DETAILED ACTION
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . The following FINAL office action is in response to Applicant communication filed on 06/04/2026 regarding application 18/654,876. Claims 1, 4-5, 15, 18-19, 29 and 32-33 have been amended. Claims 1-42 are pending and have been rejected.
Response to Amendments
2. Applicant’s amendment filed on 06/04/2026 necessitated new grounds of rejection in this office action.
Priority
3. The Examiner has noted the Applicants claiming Priority from Provisional (PRO) Application #63/499,952 filed on 05/03/2023. Therefore, the earliest effective filing date considered for this case is 05/03/2023.
Status of Claims
4. Applicant’s arguments, see page 11 of 15 filed on 06/04/2026, with respect to the Claim Objections for Claims 4-5, 18-19 and 32-33 have been fully considered, and are found to be persuasive. Therefore, the Claim Objections for Claims 4-5, 18-19 and 32-33 are withdrawn.
5. Applicant’s arguments, see pages 12-13 of 15 filed on 06/04/2026, with respect to the 35 U.S.C. § 102 (a) (1) Rejections for Claims 1, 3, 6, 8-9, 12-13, 15, 17, 20, 22-23, 26-27, 29, 31, 34, 36-37 and 40-41 have been fully considered, and are found to be not persuasive.
6. Applicant’s arguments, see page 14 of 15 filed on 06/04/2026, with respect to the 35 U.S.C. § 103 Rejections for Claims 2, 4-5, 7, 10-11, 14, 16, 18-19, 21, 23-25, 28, 30, 32-33, 35, 38-39 and 42 have been fully considered, and are found to be not persuasive. Applicant’s arguments with respect to Claims 1-42 above have been considered, but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Response to 35 U.S.C. § 101 Arguments
7. Applicant’s 35 U.S.C. § 101 arguments, filed with respect to Claims 1-42 have been fully considered, but they are found not persuasive (see Applicant Remarks, Page 12 of 15 dated 06/04/2026). Examiner respectfully disagrees.
Argument #1:
(A). Applicant argues that Claims 1-42 do not recite an abstract idea, law of nature of natural phenomenon under revised step 2a prong one of the 35 U.S.C § 101 analysis (see Applicant Remarks, Page 12 of 15, dated 06/04/2026). Examiner respectfully disagrees.
Specifically, Applicant argues that the Office Action improperly ignores multiple limitations in order to frame the claims as a mental process and also the Office Action fails to identify how the claims could possibly be “organizing human activities” under revised step 2a prong one of the 35 U.S.C § 101 analysis. Examiner respectfully disagrees.
In response, Examiner notes that the amended claims now expressly recites schedules and operation of processing facilities, but these claim limitations do not change the underlying character of Independent Claims 1, 15 and 29. Rather, they are the intended application/output of the same predictive and mathematical optimization process and do not provide a technological improvement or otherwise integrate the exception into a practical application.
Applicant's arguments have been considered but are not persuasive. The amendments do not remove the judicial exception or otherwise integrate the exception into a practical application. When Independent Claims 1, 15 and 29 for example are considered as a whole, including the newly added scheduling and facility-operation limitations, the claims remain directed to collecting information concerning an underlying system, generating predictions of future numerical values, mathematically optimizing those predicted values over multiple time periods, generating a schedule based on the optimization, and implementing the resulting schedule. The amended limitations merely specify the environment and intended use of the optimization results and do not recite a specific technological improvement.
Applicant asserts that the Office Action improperly ignores multiple limitations in framing the claims as a mental process. This is not persuasive. The § 101 analysis does not require each limitation to independently constitute an abstract idea. Rather, the claims are evaluated as a whole to determine what the claims are directed to. See Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66, 78-80 (2012); Alice Corp. v. CLS Bank International, 573 U.S. 208, 217-18 (2014); MPEP § 2106.
The Office does not disregard the limitations directed to generating optimization results;
generating schedules; operating processing facilities; or providing additional details concerning the optimizer. Instead, these limitations are considered in determining whether the claims as a whole integrates the identified abstract idea into a practical application.
The relevant inquiry is not whether the claims contain additional words describing a physical system or a computer implementation, but whether those additional elements meaningfully limit the abstract idea or instead merely apply the abstract idea to a particular technological environment.
Here, the additional limitations do not change the character of the claimed subject matter. These claims still obtains data concerning an underlying system, generates predictions of numerical values, supplies those numerical values as coefficients to an optimizer, executes the optimizer to mathematically determine optimization results, and generates schedules from those results. The newly recited operation of the processing facilities is simply the downstream implementation of the optimization result.
Thus, the Office has not ignored the amended limitations. Rather, those limitations have been considered and determined not to be sufficient to transform the abstract optimization process into patent-eligible subject matter. Applicant argues that the Office Action failed to identify what “mathematical concept” the claims recite and appears to rely on an incorrect premise that tangential association with mathematics renders claims abstract. Applicant's characterization of the rejection is not accurate. The rejection does not rest merely on the proposition that the claims are “associated with mathematics” because a computer performs the claimed steps. Rather, the claims expressly requires mathematical operations and numerical relationships as substantive steps of the claimed process.
In particular, Independent Claims 1, 15 and 29 requires: “predictions [that] represent estimated values comprising numeric values” of time-varying parameters; “at least one or more coefficient values” of terms of an optimizer; and “executing ... the optimizer to generate optimization results based on the predictions.”
These limitations do not merely involve incidental use of numbers. The numerical values and coefficients are the inputs to the claimed optimization operation, and the optimization operation produces the results upon which the schedules are based. Thus, the mathematical character of these claims arises from the claimed calculation/optimization of numerical values, not merely from the fact that a computer is used.
The USPTO's Step 2A Prong 1 guidance recognizes mathematical calculations, mathematical relationships, and mathematical formulas as abstract ideas. MPEP § 2106.04(a). The Federal Circuit likewise has repeatedly recognized claims directed to mathematical calculations or information processing as potentially abstract even when implemented using computer technology. Accordingly, the rejection is not based on the mere fact that “a computer is used.” It is based on the substantive mathematical nature of the claimed prediction and optimization process.
Claims 1, 15 and 29 require “generating ... predictions based on the obtained data” where the predictions are estimated numerical values of parameters associated with an underlying system at multiple future time periods. This limitation requires analyzing existing information and determining estimated future conditions from that information. These claims do not require a particular technological mechanism for making the prediction. It does not require, for example, a particular sensor architecture, computer architecture, data structure, unconventional prediction algorithm, or particular machine-learning architecture that improves computer functionality.
Instead, the limitation is expressed functionally in terms of the result to be obtained—predicted values. The underlying activity of reviewing information concerning a system and predicting future values based upon that information is an information-analysis activity that can constitute a mental process when considered at the level of the claimed concept. The fact that the claims require a “processing device” to perform the activity does not, by itself, remove the activity from the mental-process category. As explained by the Federal Circuit in Electric Power Group, LLC v. Alstom S.A., 830 F.3d 1350, 1353-54 (Fed. Cir. 2016), collecting information, analyzing information, and identifying information based on analysis can constitute abstract information processing even when performed using computer technology. Thus, Applicant's argument that the Office ignores the computer implementation is unpersuasive. The computer implementation is acknowledged but does not alter the abstract nature of the underlying information analysis.
The amendments concerning the optimizer likewise do not overcome the rejection. Claims 1, 15 and 29 require the predictions to comprise: “one or more coefficient values of one or more terms of the optimizer” and requires: “executing ... the optimizer to generate optimization results based on the predictions.” These limitations further demonstrate that the claimed process involves mathematical optimization. Applicant may contend that the additional optimizer details make the claims more technologically specific. However, specificity concerning the inputs, parameters, or outputs of a mathematical optimization does not necessarily transform the mathematical optimization into a technological improvement. The relevant question is whether the optimizer itself represents an improvement to computer technology or another technology, rather than merely a particular implementation of the abstract optimization.
Nothing in the amended claims require a particular unconventional computer architecture or a technological improvement in how optimization is performed. These claims specify what information is provided to the optimizer and what results are obtained, but does not establish that the optimizer itself operates in an unconventional manner that improves the functioning of the computer or the optimizer technology.
Accordingly, these amendments do not remove the claims from the mathematical-concept category.
Applicant particularly relies upon the amendments reciting: “generating one or more schedules to operate one or more processing facilities based, at least in part, on the one or more optimization results” and “causing the one or more processing facilities to operate based, at least in part, on the one or more schedules.” These limitations have been considered and do not render the claims eligible.
The scheduling limitations merely take the result of the preceding abstract optimization and use that result to establish an operating schedule. Independent Claims 1, 15 and 29 do not recite a new scheduling technology. It does not specify: a particular scheduling algorithm; a particular control architecture; a particular communications protocol; a particular actuator arrangement; a particular sensor arrangement; a particular machine configuration; or another technological mechanism that improves operation of the processing facility.
Instead, Independent Claims 1, 15 and 29 merely requires a schedule to be generated “based, at least in part, on” the optimization results and then requires the facilities to operate “based, at least in part, on” that schedule.
The limitations therefore specify the purpose and downstream use of the optimization results, rather than a technological improvement in the manner in which the facilities operate.
Applicant may argue that the amended claims are materially different because it does not merely calculate or display an optimization result, but actually causes physical processing facilities to operate. That argument is not persuasive.
The mere inclusion of a physical-world consequence does not automatically transform an otherwise abstract process into patent-eligible subject matter. The critical question is whether the physical operation represents a specific technological application of the abstract idea. Here, Independent Claims 1, 15 and 29 does not specify how the processing facilities are controlled. It merely states that they are caused to operate according to a schedule generated from the optimization results. Thus, Independent Claims 1, 15 and 29 essentially proceeds as follows: obtain information → predict values → mathematically optimize → generate schedule → operate facility according to schedule. The final physical step is the expected application of the preceding optimization. There is no requirement that the physical facility itself be technologically improved. This distinction is consistent with Electric Power Group, where the Federal Circuit explained that claims directed to collecting and analyzing information do not become eligible merely because the information concerns a physical system. 830 F.3d at 1353-54.
The reference to “processing facilities” does not meaningfully limit the abstract idea.
The claimed optimization could be applied to an underlying system by determining predicted values and generating a schedule based on those values. The processing facilities provide the environment in which the resulting schedule is used. Limiting an abstract idea to a particular technological environment does not necessarily integrate the idea into a practical application. This principle is particularly applicable here because Independent Claims 1, 15 and 29 does not identify a technological improvement specific to the processing facilities. The facilities are simply the target of the schedule produced by the optimization.
Accordingly, the recitation of processing facilities does not transform the mathematical optimization into a technological invention.
The Federal Circuit's decision in Recentive Analytics, Inc. v. Fox Corp., 134 F.4th 1205 (Fed. Cir. 2025), provides particularly relevant guidance. In Recentive, the claims involved machine-learning techniques used to generate optimized schedules. The Federal Circuit held that the claims were directed to the abstract idea of using generic machine learning to generate optimized schedules and were not saved merely because the scheduling was performed in a particular environment. The court emphasized that the claims did not claim a specific improvement to the underlying machine-learning technology. Instead, they applied generic machine-learning techniques to a particular field. The present claims are analogous.
Here, Independent Claims 1, 15 and 29: obtains information; generates predictions; provides the predictions to an optimizer; executes the optimizer; generates optimization results; generates schedules; and causes processing facilities to operate according to those schedules. Like the claims in Recentive, the claims does not identify a technological improvement in the underlying computational technology. Rather, it uses computational optimization to generate schedules for a particular operational environment. The addition of “processing facilities” therefore does not, by itself, transform the claims into a technological improvement. Recentive is particularly relevant because it rejected the proposition that applying a generic computational technique to a particular operational environment is sufficient to confer eligibility. Examiner refers Applicant to MPEP § 2106.04 (a) (2) II which states that: “the sub-groupings encompass both activity of a single person and activity that involves multiple people, and thus, certain activity between a person and a computer may fall within the "Certain Methods of Organizing Human Activities" groupings. It is noted that the number of people involved in the activity is not dispositive as to whether a claim limitation falls within this grouping. Instead, the determination should be based on whether the activity itself falls within one of the sub-groupings.”
With respect to “Mental Processes” category, Examiner refers Applicant to MPEP § 2106.04 (a) (2) (III) (C): “Claims can recite a mental process even if they are claimed as being performed on a computer. The Supreme Court recognized this in Benson, determining that a mathematical algorithm for converting binary coded decimal to pure binary within a computer’s shift register was an abstract idea. The Court concluded that the algorithm could be performed purely mentally even though the claimed procedures "can be carried out in existing computers long in use, no new machinery being necessary." 409 U.S at 67, 175 USPQ at 675. See also Mortgage Grader, 811 F.3d at 1324, 117 USPQ2d at 1699 (concluding that concept of "anonymous loan shopping" recited in a computer system claims are an abstract idea because it could be "performed by humans without a computer").” “For instance, the Examiner has reviewed Applicant’s Specification and determined that the claimed invention is described as concepts that are performed in the human mind and applicant is merely claiming that concept performed 1) on a generic computer (see Applicant’s Specification ¶ [0032]: “The processing device 202 includes any suitable number(s) and type(s) of processors or other processing devices in any suitable arrangement. Example types of processing devices 202 include one or more microprocessors, microcontrollers, reduced instruction set computers (RISCs), complex instruction set computers (CISCs), graphics processing units (GPUs), data processing units (DPUs), virtual processing units, associative process units (APUs), tensor processing units (TPUs), vision processing units (VPUs), neuromorphic chips, artificial intelligence (AI) chips, quantum processing units (QPUs), cerebras wafer-scale engines (WSEs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or discrete circuitry.”), or 2) in a computer environment (see Applicant’s Specification ¶ [0037] and Fig. 3: “An example environment 300 in which a machine learning-based approach or other approach for generating inputs for an optimizer can be used according to this disclosure. More specifically, FIG. 3 illustrates an example environment 300 for which the system 100 may be used to schedule harvesting or processing operations associated with at least one agricultural product. However, this is merely one example, and the system 100 may be used in or in conjunction with any other suitable environment.”), or 3) is merely using a computer as a tool to perform these concepts.” Thus, based on these 3 factors, Examiner maintains that the claims still recite a mental process.
In conclusion, therefore, at step 2a prong 1, Claims 1-42 are directed to the abstract idea and recited judicial exceptions under “Certain Methods of Organizing Human Activities” category or “Mental Processes” category or “Mathematical Concepts” category. Claims 1-42 are maintained as being patient ineligible under 35 U.S.C. 101 step 2a prong 1.
Argument #2:
(B). Applicant argues that Claims 1-42 recite additional elements that integrate the judicial exception into a practical application under revised step 2a prong two of the 35 U.S.C. § 101 analysis (see Applicant Remarks, Page 12 of 15, dated 06/04/2026). Examiner respectfully disagrees.
Under Step 2A Prong 2, the claims must be considered as a whole to determine whether the additional elements integrate the abstract idea into a practical application. Here, for example, Independent Claims 1, 15 and 29 do not. The additional elements do not: improve the functioning of the computer; improve the optimizer itself; improve the prediction technology; improve a particular data-processing technique; improve the processing facilities; improve a particular control system; or effect a particular technological transformation.
Instead, the claims apply mathematical prediction and optimization to an underlying system and uses the resulting optimization to generate a schedule. These claims therefore do not recite a technological solution to a technological problem. It recites a solution to an operational optimization problem using mathematical information processing. That distinction is critical.
Applicant may rely on the processing device and processing facilities as physical components. However, simply naming a processor or physical facility is insufficient. The relevant question is whether the claimed machine is integral to the claimed method in a manner that meaningfully limits the exception, rather than merely serving as a generic tool for performing the abstract calculations. The “processing device” performs generic functions: obtaining data; generating predictions; providing values; executing the optimizer; and generating results. The “processing facilities” merely operate according to the resulting schedule. Neither is claimed with sufficient technological specificity to demonstrate an improvement in the functioning of the machine or facility.
Applicant may further argue that operating processing facilities necessarily produces a physical result. However, the claims do not specify a particular transformation of an article or material. These claims merely cause unspecified processing facilities to operate according to an unspecified schedule. Accordingly, the physical operation is too general to establish a particular technological transformation.
The amended limitations essentially require the abstract optimization to be put into practice. But “putting an abstract idea into practice” in the ordinary sense is not sufficient. The claim must apply the exception in a manner that imposes meaningful limitations and reflects a technological application. Here, the schedule is simply the output of the optimization, and operating the facility according to that schedule is simply the intended use of the output. There is no additional technical mechanism that changes the character of the claim.
Independent Claims 1, 15 and 29: With respect to reliance on (e.g., “an optimizer” & “underlying system”) as additional elements when considered individually and as an ordered combination (as a whole) for the claim limitations for Independent Claims 1, 15 and 29, these additional elements do not provide limitations that are indicative of integration into a practical application due to: (1) reciting mere instructions to implement an abstract idea on a computer or using a computer as a tool to “apply” the recited judicial exceptions (see MPEP § 2106.05(f)) or (2) limiting to a particular field of use or technological environment pertaining to providing predictions to an optimizer whereby the optimizer generates optimization results based on the predictions association with an underlying system using a computer in an agricultural environment (see MPEP § 2106.05 (h)).
Examiner notes that these limitations utilize generic computing devices ("at least one processing device") to perform data obtaining, predicting, optimizing, and scheduling. The final step ("causing the one or more processing facilities to operate") purports to tie the math to physical machinery (processing facilities). However, since the operation of the facility merely carries out the abstract schedule using operation commands without improving how the facility or the processing equipment itself operates technologically, it fails to integrate the abstract idea into a practical application under current USPTO guidance and Federal Circuit precedent (e.g., Recurrent / generic application of AI/optimization). Because the steps recite generic data gathering, calculation, and control without specifying an improvement to the computational machinery or the physical facility's internal operation, these claims are not integrated into a practical application.
Additionally, certain/particular claim limitations in Independent Claims 1, 15 and 29 recite steps of “receiving data” (e.g., “obtaining, using at least one processing device, data from one or more data sources, wherein the data is associated with or affects an underlying system to be optimized”) when evaluated as additional elements, these activities at most amount to insignificant extra-solution activities (see MPEP § 2106.05 (g)), which have been recognized as Well-Understood, Routine and Conventional (WURC), and thus insufficient to add significantly more to the abstract idea. See MPEP § 2106.05(d) ii - Receiving or Transmitting Data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network).
Additionally, from MPEP § 2106.04 (d): “It is notable that mere physicality or tangibility of an additional element or elements is not a relevant consideration in Step 2A Prong Two.” As the Supreme Court explained in Alice Corp., mere physical or tangible implementation of an exception does not guarantee eligibility. Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 573 U.S. 208, 224, 110 USPQ2d 1976, 1983-84 (2014) ("The fact that a computer ‘necessarily exist[s] in the physical, rather than purely conceptual, realm,’ is beside the point"). See also Genetic Technologies Ltd. v. Merial LLC, 818 F.3d 1369, 1377, 118 USPQ2d 1541, 1547 (Fed. Cir. 2016) (steps of DNA amplification and analysis are not "sufficient" to render claim 1 patent eligible merely because they are physical steps). Conversely, the presence of a non-physical or intangible additional element does not doom the claims, because tangibility is not necessary for eligibility under the Alice/Mayo test. Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 118 USPQ2d 1684 (Fed. Cir. 2016) ("that the improvement is not defined by reference to ‘physical’ components does not doom the claims"). See also McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1315, 120 USPQ2d 1091, 1102 (Fed. Cir. 2016), (holding that a process producing an intangible result (a sequence of synchronized, animated characters) was eligible because it improved an existing technological process).
In addition, these limitations fail to provide an improvement to the functioning of a computer or to any other technology or technical field, fail to apply the exception with a particular machine, fail to apply the judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, fail to effect a transformation of a particular article to a different state or thing, and fail to apply/use the abstract idea in a meaningful way beyond generally linking the use of the judicial exception to a particular technological environment. Accordingly, because the Step 2A Prong One and Prong Two analysis resulted in the conclusion that the claims are directed to an abstract idea, additional analysis under Step 2B of the eligibility inquiry must be conducted in order to determine whether any claim element or combination of elements amount to significantly more than the judicial exception.
Therefore, in conclusion, Examiner maintains that Claims 1-42 do not recite additional elements that integrate the judicial exception into a practical application under step 2a prong 2 of the 35 U.S.C. § 101 analysis. Claims 1-42 are patent ineligible under step 2a prong 2 of the 35 U.S.C. § 101 analysis.
Argument #3:
(C). Applicant argues that Claims 1-42 recite additional elements that amount to significantly more than the recited judicial exceptions under revised step 2B of the 35 U.S.C. § 101 analysis (see Applicant Remarks, Page 12 of 15, dated 06/04/2026). Examiner respectfully disagrees.
In response, Examiner refers Applicant to Examiner’s 35 U.S.C. 101 analysis section (e.g., Claim Rejections - 35 U.S.C. § 101 section shown below) shown for step 2B particularly for Independent Claims 1, 15 and 29. The claims do not recite additional elements that amount to significantly more than the recited judicial exceptions, because they are merely directed to the particulars of the abstract idea and likewise do not add significantly more to the above-identified judicial exceptions. The limitations are directed to limitations referenced in MPEP § 2106.05I.A. that are not enough to qualify as significantly more when recited in these claims with the abstract idea which include: (1) adding the words “apply it” (or an equivalent) with the judicial exception, (2) or mere instructions to implement an abstract idea on a computer and providing the results to the user on a computer, and (3) generally linking the use of the judicial exception to a particular technological environment or field of use.
Independent Claims 1, 15 and 29: With respect to reliance on (e.g., “an optimizer” & “underlying system”) as additional elements when considered individually and as an ordered combination (as a whole) for the claim limitations for Independent Claims 1, 15 and 29, these additional elements do not amount to significantly more than the judicial exceptions under step 2B due to: (1) reciting mere instructions to implement an abstract idea on a computer or using a computer as a tool to “apply” the recited judicial exceptions (see MPEP § 2106.05(f)) or (2) limiting to a particular field of use or technological environment pertaining to providing predictions to an optimizer whereby the optimizer generates optimization results based on the predictions association with an underlying system using a computer in an agricultural environment (see MPEP § 2106.05 (h)). Additionally, certain/particular claim limitations in Independent Claims 1, 15 and 29 recite steps of “receiving data” (e.g., “obtaining, using at least one processing device, data from one or more data sources, wherein the data is associated with or affects an underlying system to be optimized”) when evaluated as additional elements, these activities at most amount to insignificant extra-solution activities (see MPEP § 2106.05 (g)), which have been recognized as Well-Understood, Routine and Conventional (WURC), and thus insufficient to add significantly more to the abstract idea. See MPEP § 2106.05(d) ii - Receiving or Transmitting Data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). The components recited: "at least one processing device," "data sources," "optimizer," and "processing facilities" are all claimed at a high level of functionality. They describe hardware components performing their typical, expected functions (storing, calculating, communicating, and running optimization software). The ordered combination: The sequence (gather data -> predict/estimate values -> feed coefficients to an optimizer -> computer optimization -> output schedule -> run facility) outlines a mental/mathematical workflow adapted to a computer system. There is no specific, unconventional technological improvement to the computer's architecture, no specialized machine-learning training breakthrough specified in the steps, and no novel physical control mechanism for the facilities. The claims do not amount to significantly more than the abstract mathematical/mental process of predicting, optimizing, and scheduling. Under 35 U.S.C. 101, the claims are prima facie ineligible.
Moreover, Examiner refers Applicant to BSG Tech LLC v. Buyseasons Inc. decision (Aug. 15, 2018) court case noting that: “But the relevant inquiry is not whether the claimed invention as a whole is unconventional or non-routine. At Step two, we “search for an ‘inventive concept’… that is sufficient to ensure that the patent in practice amounts to significantly more than a patent upon the [ineligible concept] itself.” Alice, 134 S. Ct. at 2355 (internal quotation marks omitted) (quoting Mayo, 566 U.S. at 72-73). But this simply restates what we have already determined is an abstract idea. At Alice step two, it is irrelevant whether considering historical usage information while inputting data may have been non-routine or unconventional as a factual matter. As a matter of law, narrowing or reformulating an abstract idea does not add “significantly more” to it. See SAP Am., Inc. v. InvestPic, LLC. No. 2017-2081, slip op. at 14 (Fed. Cir. 2018). Applicant’s suggestion that specific limitations (or the claimed invention as a whole) must be shown to be well-understood, routine, and conventional to support the conclusion of subject matter ineligibility is not persuasive.
Even assuming, arguendo, that Independent Claims 1, 15 and 29 includes additional elements sufficient to proceed to Step 2B, those elements do not amount to significantly more than the abstract idea. The individual elements are conventional computer and operational functions: obtaining data; processing data; generating predictions; performing numerical optimization; generating schedules; and operating equipment according to a schedule. The claims do not identify an unconventional manner of performing these functions. Nor does the ordered combination produce an unconventional technological result. The ordered combination simply performs the ordinary sequence expected when using optimization results to schedule facility operations.
Applicant's amendments should also be considered collectively. The relevant combination is: obtaining system data → generating predictions → supplying numerical coefficients → executing optimization → generating optimization results → generating schedules → operating facilities according to schedules. Nothing in this combination changes the underlying nature of the claim. The mathematical optimization remains the central activity. The additional scheduling and facility-operation limitations merely provide the intended downstream use of the optimization results. Consequently, even when considered as an ordered combination, the limitations do not provide an inventive concept. This is consistent with the Federal Circuit's reasoning in Electric Power Group and Recentive, where generic computer implementation and application of abstract information processing to a particular operational environment were insufficient to supply an inventive concept. Applicant asserts that the Office Action failed to establish a prima facie case of ineligibility.
This assertion is not persuasive. The Office has identified the specific claim limitations that constitute the abstract concept and explained why the additional limitations do not integrate the concept into a practical application. Moreover, an eligibility analysis is not defeated merely because Applicant disagrees with the characterization of the abstract idea. The proper inquiry is whether the claims, viewed as a whole, are directed to a judicial exception and, if so, whether additional elements amount to significantly more. Here, the claim's substantive steps remain directed to information gathering, prediction, mathematical optimization, and scheduling. The amended physical implementation does not change that characterization. Accordingly, the Office maintains that a prima facie case of ineligibility has been established.
Applicant's contention that the § 101 rejection is “moot” because of the amendments are also not persuasive. An amendment does not automatically moot a pending § 101 rejection merely because it adds additional limitations. The amended claims must itself be analyzed under the applicable § 101 framework. Here, the amendments were specifically considered. They do not remove the abstract idea because the newly added limitations merely: use the optimization results to generate schedules; and operate the processing facilities according to those schedules. These limitations do not claim a specific technological improvement.
The ordered combination of elements in the Dependent Claims (including the limitations inherited from the parent claim(s)) add nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Accordingly, the subject matter encompassed by the dependent claims fails to amount to a practical application or significantly more than the abstract idea itself. Therefore, under Step 2B, Claims 1-42 do not include additional elements that are sufficient to amount to significantly more than the recited judicial exceptions.
In summary, for at least the foregoing reasons, Applicant's arguments are not persuasive. The amended claim remains directed to the abstract concept of obtaining information concerning an underlying system, generating predicted numerical values for future time periods, mathematically optimizing the predicted values, generating an operating schedule based on the optimization, and implementing the resulting schedule. The limitations directed to the optimizer recite mathematical values and optimization operations rather than an improvement to computer technology or optimization technology. The newly added limitations directed to generating schedules and causing processing facilities to operate according to the schedules merely apply the optimization results to a particular operational environment and do not recite a specific technological improvement to the processing facilities or a particular control mechanism. Thus, the additional limitations do not integrate the judicial exception into a practical application under Step 2A Prong 2. Further, the processing device, data acquisition, prediction, optimization, scheduling, and facility-operation functions, considered individually and as an ordered combination, do not amount to significantly more than the abstract idea under Step 2B. Accordingly, the amendments do not overcome the § 101 rejection, and the rejection is maintained. Thus, Claims 1-42 are ineligible with respect to the 35 U.S.C. § 101 analysis.
Claim Rejections - 35 USC § 101
8. 35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
9. Claims 1-42 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1-42 are each focused to a statutory category namely, a “method” or a “process” (Claims 1-14), an “apparatus” or a “system” (Claims 15-28) and a “non-transitory computer readable medium” or an “article of manufacture” (Claims 29-42).
Step 2A Prong One: Independent Claims 1, 15 and 29 recites limitations that set forth the abstract idea(s), namely (see in bold except where strikethrough):
“obtaining, , data , wherein the data is associated with or affects to be optimized” (see Independent Claim 1);
“generating, , predictions based on the obtained data, wherein the predictions represent estimated values comprising numeric values of one or more time-varying parameters associated with at a plurality of time periods in an optimization horizon” (see Independent Claim 1);
“providing, , the predictions wherein the predictions comprise at least one or more coefficient values of one or more terms” (see Independent Claim 1);
“executing, , to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon
“generating one or more schedules to operate one or more processing facilities based, at least in part, on the optimization results” (see Independent Claim 1);
“causing the one or more processing facilities to operate based, at least in part, on the one or more schedules” (see Independent Claim 1);
“” (see Independent Claim 15);
“” (see Independent Claim 29);
“obtain data , wherein the data is associated with or affects to be optimized” (see Independent Claims 15 and 29);
“generate predictions based on the obtained data, wherein the predictions represent estimated values comprising numeric values of one or more time-varying parameters associated with at a plurality of time periods in an optimization horizon” (see Independent Claims 15 and 29);
“provide the predictions wherein the predictions comprise at least one or more coefficient values of one or more terms” (see Independent Claims 15 and 29);
“execute to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon
“generate one or more schedules to operate one or more processing facilities based, at least in part, on the optimization results” (see Independent Claims 15 and 29);
“cause the one or more processing facilities to operate based, at least in part, on the one or more schedules” (see Independent Claims 15 and 29).
Here, for Independent Claims 1, 15 and 29, these steps recite an abstract idea directed to collecting data, mathematically generating predictions and optimization coefficients, running an optimization algorithm, and using the resulting production schedule to control physical operations in an agricultural harvesting of crops in growing areas/lots. This overall abstract concept groups cleanly into: Mental Processes: Steps involving mathematical calculations, generating predictions, estimating numeric values, and executing optimizers/algorithms that can conceptually be performed in the human mind or via generic pencil-and-paper math. Certain Methods of Organizing Human Activities: Steps involving generating business/operational schedules and planning facility management.
Examiner analyzes each claim limitation step both individually and as whole for Independent Claims 1, 15 and 29 below. For example; the step of “Obtaining data from one or more data sources…” recites the concept of data acquisition/collection. This is classified and falls under the Certain Methods of Organizing Human Activities (managing information/commercial or operational data collection) grouping/category or Mental Processes (receiving/noting facts) grouping/category. Standing alone or as part of data acquisition, receiving or gathering information is a precursor to analysis and falls under basic information collection or mental receipt of data. Secondly, the step of “Generating predictions based on the obtained data (numeric values of time-varying parameters)…” recites the concept of mathematical calculation, estimation, and predictive modeling. This is classified and falls under the Mental Processes (evaluating, calculating, and estimating values or outcomes). This step recites calculating, estimating, or predicting numeric values, which falls squarely under mathematical concepts and mental processes that can be performed in the human mind or via basic mathematical algorithms. Thirdly, the step of “Providing the predictions to an optimizer (coefficient values of terms)…” recites the concept of formatting or communicating data/mathematical values for algorithmic use. This is classified and falls under Mental Processes / Certain Methods of Organizing Human Activities (rules or formats for organizing inputs). Transmitting calculated parameters or formatting them as mathematical coefficients for an equation is a mental/mathematical preparation step. Fourth, the step of “Executing the optimizer to generate optimization results…” recites mathematical optimization and running formulas/algorithms. classified and falls under Mental Processes (mathematical calculation and mathematical operations). Running an optimization algorithm or mathematical model to find optimal values is a pure mathematical calculation and mental process. Fifth, the step of “Generating one or more schedules to operate processing facilities…” recites the concept of creating a plan, agenda, or schedule. This classified and falls under the Certain Methods of Organizing Human Activities (managing, establishing, or tracking a schedule/business operational plan). Creating a schedule or a plan of action based on calculated results represents a method of organizing human or facility activity. Lastly, the step of “Causing the processing facilities to operate based on the schedules…” describes that while control actions can edge toward concrete utility, the causation based on a schedule at this stage acts as post-solution activity or generic application of a calculated result. This is classified and falls under being functionally tied to Certain Methods of Organizing Human Activities (implementing an organizational plan) executed via automation. The physical action here ("causing to operate" via generic control signals) executes the previously derived abstract schedule.
Therefore, these abstract idea limitations (as identified above in bold), under their broadest reasonable interpretation of the claims as a whole, cover performance of their limitations as “Certain Methods of Organizing Human Activities” which pertains to (1) managing personal behavior or relationships or interactions between people (including teachings or following rules or instructions).
Additionally, or alternatively, these abstract idea limitations (as identified above in bold), under the broadest reasonable interpretation of the claims as a whole, cover performance of their limitations as “Mental Processes” which pertains to (2) concepts performed in the human mind (including observations or evaluations or judgments) or (3) using pen and paper as a physical aid, in order to help perform these mental steps does not negate the mental nature of these limitations. The use of "physical aids" in implementing the abstract mental process, does not preclude these claims from reciting an abstract idea. See MPEP § 2106.04(a) III C. Additionally, or alternatively, these abstract idea limitations (as identified above in bold), under the broadest reasonable interpretation of the claims as a whole, cover performance of their limitations as “Mathematical Concepts” which pertains to (4) mathematical relationships.
That is, other than reciting the additional elements of (e.g., “one or more data sources” & “an optimizer” & “underlying system” & “one or more processors” & “at least one processing device”, etc…), nothing in the claim elements precludes the steps from being performed as “Mental Processes” which pertains to (1) concepts performed in the human mind (including observations or evaluations or judgments or opinions) or (2) using pen and paper as a physical aid and additionally or alternatively as “Certain Methods of Organizing Human Activities” which pertains to (3) managing personal behavior or relationships or interactions between people (including teachings or following rules or instructions) and additionally or alternatively as “Mathematical Concepts” which pertains to (4) mathematical relationships.
Moreover, the mere recitation of generic computer components such as (e.g., “one or more processors” & “at least one processing device”) does not take the claims out of “Certain Methods of Organizing Human Activities” or “Mental Processes” or “Mathematical Concepts” Groupings.
Therefore, at step 2a prong 1, Yes, Claims 1-42 recite an abstract idea. We proceed onto analyzing the claims at step 2a prong 2.
Step 2A Prong Two: With respect to Step 2A Prong Two of the eligibility inquiry (as explained in MPEP § 2106.04(d)), the judicial exception is not integrated into a practical application. Independent Claims 1 and 15 recites additional elements directed to: (e.g., “at least one processing device” & “one or more data sources”). Independent Claim 29 recites additional elements directed to: (e.g., “one or more processors” & “one or more data sources”). These additional elements have been considered individually and in combination, but fail to integrate the abstract idea into a practical application because they amount to using generic computing elements or instructions (software) to perform the abstract idea, similar to adding the words “apply it” (or an equivalent), which merely serves to link the use of the judicial exception to a particular technological environment. See MPEP § 2106.05(f) and MPEP § 2106.05(h).
Independent Claims 1, 15 and 29: With respect to reliance on (e.g., “an optimizer” & “underlying system”) as additional elements when considered individually and as an ordered combination (as a whole) for the claim limitations for Independent Claims 1, 15 and 29, these additional elements do not provide limitations that are indicative of integration into a practical application due to: (1) reciting mere instructions to implement an abstract idea on a computer or using a computer as a tool to “apply” the recited judicial exceptions (see MPEP § 2106.05(f)) or (2) limiting to a particular field of use or technological environment pertaining to providing predictions to an optimizer whereby the optimizer generates optimization results based on the predictions association with an underlying system using a computer in an agricultural environment (see MPEP § 2106.05 (h)).
Examiner notes that these limitations utilize generic computing devices ("at least one processing device") to perform data obtaining, predicting, optimizing, and scheduling. The final step ("causing the one or more processing facilities to operate") purports to tie the math to physical machinery (processing facilities). However, since the operation of the facility merely carries out the abstract schedule using operation commands without improving how the facility or the processing equipment itself operates technologically, it fails to integrate the abstract idea into a practical application under current USPTO guidance and Federal Circuit precedent (e.g., Recurrent / generic application of AI/optimization). Because the steps recite generic data gathering, calculation, and control without specifying an improvement to the computational machinery or the physical facility's internal operation, these claims are not integrated into a practical application.
In addition, these limitations fail to provide an improvement to the functioning of a computer or to any other technology or technical field, fail to apply the exception with a particular machine, fail to apply the judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, fail to effect a transformation of a particular article to a different state or thing, and fail to apply/use the abstract idea in a meaningful way beyond generally linking the use of the judicial exception to a particular technological environment.
Accordingly, because the Step 2A Prong One and Prong Two analysis resulted in the conclusion that the claims are directed to an abstract idea, additional analysis under Step 2B of the eligibility inquiry must be conducted in order to determine whether any claim element or combination of elements amount to significantly more than the judicial exception. Therefore, at step 2a prong 2, Claims 1-42 are directed to the abstract idea and do not recite additional elements that integrate into a practical application.
Step 2B: (As explained in MPEP § 2106.05), it has been determined that the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. Independent Claims 1 and 15 recites additional elements directed to: (e.g., “at least one processing device” & “one or more data sources”). Independent Claim 29 recites additional elements directed to: (e.g., “one or more processors” & “one or more data sources”). These elements have been considered individually and in combination, but fail to add significantly more to the claims because they amount to using computing elements or instructions (software) to perform the abstract idea, similar to adding the words “apply it” (or an equivalent), which merely serves to link the use of the judicial exception to a particular technological environment (computing environment) and does not amount to significantly more than the abstract idea itself. See MPEP § 2106.05 (h) and See MPEP § 2106.05 (f). Notably, Applicant’s Specification suggests that the claimed invention relies on nothing more than a general-purpose computer executing the instructions to implement the invention (see at least Applicant’s Specification ¶ [0032]: “The processing device 202 includes any suitable number(s) and type(s) of processors or other processing devices in any suitable arrangement. Example types of processing devices 202 include one or more microprocessors, microcontrollers, reduced instruction set computers (RISCs), complex instruction set computers (CISCs), graphics processing units (GPUs), data processing units (DPUs), virtual processing units, associative process units (APUs), tensor processing units (TPUs), vision processing units (VPUs), neuromorphic chips, artificial intelligence (AI) chips, quantum processing units (QPUs), cerebras wafer-scale engines (WSEs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or discrete circuitry.”).
Independent Claims 1, 15 and 29: With respect to reliance on (e.g., “an optimizer” & “underlying system”) as additional elements when considered individually and as an ordered combination (as a whole) for the claim limitations for Independent Claims 1, 15 and 29, these additional elements do not amount to significantly more than the judicial exceptions under step 2B due to: (1) reciting mere instructions to implement an abstract idea on a computer or using a computer as a tool to “apply” the recited judicial exceptions (see MPEP § 2106.05(f)) or (2) limiting to a particular field of use or technological environment pertaining to providing predictions to an optimizer whereby the optimizer generates optimization results based on the predictions association with an underlying system using a computer in an agricultural environment (see MPEP § 2106.05 (h)). Additionally, certain/particular claim limitations in Independent Claims 1, 15 and 29 recite steps of “receiving data” (e.g., “obtaining, using at least one processing device, data from one or more data sources, wherein the data is associated with or affects an underlying system to be optimized”) when evaluated as additional elements, these activities at most amount to insignificant extra-solution activities (see MPEP § 2106.05 (g)), which have been recognized as Well-Understood, Routine and Conventional (WURC), and thus insufficient to add significantly more to the abstract idea. See MPEP § 2106.05(d) ii - Receiving or Transmitting Data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). The components recited: "at least one processing device," "data sources," "optimizer," and "processing facilities" are all claimed at a high level of functionality. They describe hardware components performing their typical, expected functions (storing, calculating, communicating, and running optimization software). The ordered combination: The sequence (gather data -> predict/estimate values -> feed coefficients to an optimizer -> computer optimization -> output schedule -> run facility) outlines a mental/mathematical workflow adapted to a computer system. There is no specific, unconventional technological improvement to the computer's architecture, no specialized machine-learning training breakthrough specified in the steps, and no novel physical control mechanism for the facilities. The claims do not amount to significantly more than the abstract mathematical/mental process of predicting, optimizing, and scheduling. Under 35 U.S.C. 101, the claims are prima facie ineligible.
In addition, when taken as an ordered combination, the ordered combination adds nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements integrates the abstract idea into a practical application. Therefore, when viewed as a whole, these additional claim elements do not provide meaningful limitations to transform the abstract idea into a practical application of the abstract idea or that, as an ordered combination, amount to significantly more than the abstract idea itself.
Dependent Claims 2-14, 16-28 and 30-42 recite additional elements directed to: (e.g., “agricultural system” & “processing facility” & “reinforcement learning” & “Monte Carlo simulations” & “equipment”), and when considered individually and as an ordered combination (as a whole) with the limitations recite the same abstract idea(s) as shown in Independent Claims 1, 15 and 29 along with further steps/details that could be performed as “Mental Processes” which pertains to (1) concepts performed in the human mind (including observations or evaluations or judgments) or (2) using pen and paper as a physical aid and additionally or alternatively as “Certain Methods of Organizing Human Activities” which pertains to (3) managing personal behavior (including teachings or following rules or instructions) and additionally or alternatively as “Mathematical Concepts” which pertains to (4) mathematical relationships.
Dependent Claims 2-3, 5-6, 10, 12, 14, 16-17, 19-20, 24, 26, 28, 30-31, 33-34, 38, 40 and 42 further narrow the abstract ideas, and are therefore still ineligible for the reasons previously provided in Steps 2A Prong 2 and 2B for Independent Claims 1, 15 and 29. Dependent Claims 4, 7, 11, 18, 21, 25, 32, 35 and 39: With respect to reliance on (e.g., “agricultural system” (see Dependent Claims 4, 18 and 32) & “processing facility” (see Dependent Claims 7, 21 and 35) & “reinforcement learning” & “Monte Carlo Simulations” (see Dependent Claims 11, 25 and 39)) as additional elements shown in Dependent Claims 4, 7, 11, 18, 21, 25, 32, 35 and 39 when considered individually and as an ordered combination (as a whole) in view of these claim limitations, these additional elements do not provide limitations that are indicative of integration into a practical application under step 2a prong 2 and also do not recite additional elements that amount to significantly more than the recited judicial exceptions under step 2B due to: (1) recites mere instructions to implement an abstract idea on a computer or using a computer as a tool to “apply” the recited judicial exceptions by providing the results to the user on a computer (see MPEP § 2106.05 (f)) or (2) limiting to a particular field of use or technological environment pertaining to providing predictions to an optimizer whereby the optimizer generates optimization results based on the predictions association with an underlying system using a computer in an agricultural environment or an agricultural field of use (see MPEP § 2106.05 (h)).
Dependent Claims 8-9, 13, 22-23, 27, 36-37 and 41: With respect to reliance on (e.g., “machine learning model”) as additional elements shown in Dependent Claims 8-9, 13, 22-23, 27, 36-37 and 41 when considered individually and as an ordered combination (as a whole) in view of these claim limitations, these additional elements do not provide limitations that are indicative of integration into a practical application under step 2a prong 2 and also do not recite additional elements that amount to significantly more than the recited judicial exceptions under step 2B due to: limiting to a particular field of use or technological environment pertaining to training to generated the predictions when obtained data includes imperfections and using the feedback to compensate for prediction errors associated with the predictions using a computer in an agricultural environment or an agricultural field of use (see MPEP § 2106.05 (h)).
The additional element of “machine learning” or “machine learning model” in certain/particular claims does not amount to significantly more than the judicial exception under step 2B due to being expressly recognized as Well-Understood, Routine and Conventional (WURC) in the art. See for example; US PG Pub (US 2020/0090107 A1) hereinafter McKeeman. McKeeman at ¶ [0163]: “Intelligent matching agent 290 may be based on an artificial intelligence model, neural network, or other machine learning algorithm which may, in examples, be trained to recognize and identify compatibility between information fields comprised within producer records, equipment owner records and operator records held within storages 201, 202 and 203.” See for example; US PG Pub (US 2020/0272971 A1) hereinafter Ruff, et. al. See Ruff at ¶ [0225]: “Other machine-learning methods known to someone skilled in the art for capturing various relationships between the seeding rate (in conjunction with other attributes) and the crop yield lift, such as neural networks or regression techniques, can also be used.” See Ruff at ¶ [0347]: “The model may be a regression model, such as a generalized additive model (GAM), a tree-based model, a machine learning model, and/or a neural network model. The model may be configured to estimate a distribution, such as a sinh-arcsinh (SHASH) distribution. Alternatively, the agricultural intelligence computer system may use alternative methods of quantifying uncertainty, such as Monte Carlo sampling.”
The ordered combination of elements in the Dependent Claims (including the limitations inherited from the parent claim(s)) add nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Accordingly, the subject matter encompassed by the dependent claims fails to amount to a practical application or significantly more than the abstract idea itself. Therefore, under Step 2B, Claims 1-42 do not include additional elements that are sufficient to amount to significantly more than the recited judicial exceptions. Thus, Claims 1-42 are ineligible with respect to the 35 U.S.C. § 101 analysis.
Claim Rejections - 35 USC § 103
10. This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
11. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
12. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
13. Claims 1, 3-4, 6-9, 12-13, 15, 17-18, 20-23, 26-27, 29, 31-32, 34-37 and 40-41 are rejected under 35 U.S.C. 103 as being unpatentable over US PG Pub (US 2019/0057461 A1) hereinafter Ruff, et. al, in view of NPL Document: "Scheduling planting time through developing an optimization model and analysis of time series growing degree units." (2022), hereinafter Ansarifar, et. al., and in further view of US PG Pub (US 2009/0234695 A1) hereinafter Kapadi, et. al.
Regarding Independent Claim 1, Ruff method for a machine learning-based production optimizers teaches the following:
- obtaining (see at least Ruff: ¶ [0083] & ¶ [0091] & ¶ [0198]. Ruff teaches that the system 130 may obtain or ingest data under user 102 control, on a mass basis from a large number of growers who have contributed data to a shared database system. This form of obtaining data may be termed “manual data ingest” as one or more user-controlled computer operations are requested or triggered to obtain data for use by the system 130. See also Ruff at ¶ [0083]: The nitrogen map may include one or more user input features, such as dials or slider bars, to dynamically change the nitrogen planting and practices programs so that a user may optimize his nitrogen map, such as to obtain a preferred amount of surplus to shortfall. The user may then use his optimized nitrogen map and the related nitrogen planting and practices programs to implement one or more scripts, including variable rate (VR) fertility scripts. Similar instructions to the nitrogen instructions 210 could be used for application of other nutrients (such as phosphorus and potassium), application of pesticide, and irrigation programs. See also Ruff at ¶ [0198]: The system 130 is programmed to present summaries, tips, or further recommendations generated from analyzing the data obtained from the multitude of prescribed experiments across grower fields.), using at least one processing device (see at least Ruff: ¶ [0114-0116] & Fig. 1 & Fig. 4.), data from one or more data sources (see at least Ruff: ¶ [0107-0108] & ¶ [0201]. Ruff teaches that the agricultural intelligence computer system 130 is configured or programmed to implement agronomic data preprocessing of field data received from one or more data sources. The field data received from one or more data sources may be preprocessed for the purpose of removing noise, distorting effects, and confounding factors within the agronomic data including measured outliers that could adversely affect received field data values. The historical data can be obtained from internal trials and experiments or from external data sources. The number of fields can have common values in certain characteristics, such as the crop hybrid grown in a field, the location of a field, or the yield lift management practice for a field.), wherein the data is associated with or affects an underlying system to be optimized (see at least Ruff: ¶ [0105-0107] & ¶ [0109] & ¶ [0201]. Ruff teaches that the agricultural intelligence computer system 130 is programmed or configured to create an agronomic model. In this context, an agronomic model is a data structure in memory of the agricultural intelligence computer system 130 that comprises field data 106, such as identification data and harvest data for one or more fields. The agronomic model may also comprise calculated agronomic properties which describe either conditions which may affect the growth of one or more crops on a field, or properties of the one or more crops, or both. seeds and planting instructions 208 are programmed to provide tools for seed selection, hybrid placement, and script creation, including variable rate (VR) script creation, based upon scientific models and empirical data. This enables growers to maximize yield or return on investment through optimized seed purchase, placement and population. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method.);
- generating (see at least Ruff: ¶ [0194-0197] & ¶ [0201] & ¶ [0208-0211].), using the at least one processing device (see at least Ruff: ¶ [0114-0116] & Fig. 1 & Fig. 4.), predictions based on the obtained data, wherein the predictions represent estimated values comprising numeric values of one or more time-varying parameters (see at least Ruff: ¶ [0105-0106] & ¶ [0212-0123]. Ruff teaches that the system 130 is programmed to further analyze the data, to adjust the predictions or the plans for the prescribed experiments, or to glean specific insight that can be used in designing future experiments. Such analysis can be performed periodically, at the end of a season or a year, or upon request by a grower. When a prescribed experiment was not properly carried out, the predicted result might not be obtained, and the system 130 can be programmed to adjust the prediction based on how the plan for the prescribed experiment was followed. For example, the system 130 can be configured to consider that the actual seeding rate was only 80% of the prescribed seeding rate overall, due to erroneous calibration of the agricultural implement, the skipping of certain planting steps, or other reasons, in determining the predicted crop yield might be only 80% of or otherwise less than the predicted or recommended crop yield. The system 130 can also be programmed to generate a series of remedial steps in order to realize the original prediction. For example, when the actual seeding rate was only 80% of the prescribed seeding rate overall, the system 130 can be configured to compensate for it by prescribing a seeding rate that was 20% or otherwise higher than originally prescribed for the rest of the experiment. See also Ruff at ¶ [0105]: The agronomic factors may also be used to estimate one or more crop related results, such as agronomic yield. The agronomic yield of a crop is an estimate of quantity of the crop that is produced, or in some examples the revenue or profit obtained from the produced crop. See also Ruff at ¶ [0211]: The group of fields may be selected from those fields that share values with the grower's field in certain characteristics, such as the crop hybrid grown in a field, the predicted yield lift for a change in management practice for a field, or the location of a field. See also Ruff at ¶ [0242]: The agricultural intelligence computing system may additionally alter the predicted results of the trial based on identified modifications to the testing locations. For example, the agricultural intelligence computing system may predict an increase in yield of 30 bushels/acre for an application of 40 lbs/acre of nitrogen. If the agricultural intelligence computing system detects that only 30 lbs/acre of nitrogen has been applied to a field, the agricultural intelligence computing system may lower the predicted increase in yield of 30 bushels/acre.).) associated with the underlying system at a plurality of time periods in an optimization horizon (see at least Ruff: ¶ [0185] & ¶ [0191-0194] & ¶ [0211]. Ruff teaches that the objective is to enable proper implementation of the prescribed experiments in order to achieve the predicted results. For the variables involved in the scheme, the system 130 is programmed to compare the actual value, such as the volume of seeds actually planted at a specific location within a particular period of time, such as one hour, and the prescribed value. See also Ruff at ¶ [0194]: The system 130 is programmed to further analyze the data, to adjust the predictions or the plans for the prescribed experiments, or to glean specific insight that can be used in designing future experiments. Such analysis can be performed periodically, at the end of a season or a year, or upon request by a grower. See also Ruff at ¶ [0211]: In step 1702, the system 130 is programmed to receive crop seeding rate data and corresponding crop yield data over a period of time regarding a group of fields associated with a plurality of grower devices. Such data is used to establish benchmarks for determining product responsiveness to a seeding rate increase for a grower's field. The group of fields may be selected from those fields that share values with the grower's field in certain characteristics, such as the crop hybrid grown in a field, the predicted yield lift for a change in management practice for a field, or the location of a field.)
Regarding Independent Claim 1, Ruff method for machine learning-based production optimizers does not explicitly disclose, but Ansarifar, et. al. in the analogous art for the method machine learning-based production optimizers does disclose the following:
- providing, using the at least one processing device, the predictions to an optimizer (see at least Ansarifar: Figs. 7-8 & Pages 7-8 under “Experiment Settings”. Ansarifar teaches that to tune the parameter, we used a time-wise five-fold cross-validation, as shown in Figure 6. Each fold corresponds to particular six months as test data for prediction, and the rest of the data from 2009 up to test data corresponds to the training set. We applied Adam optimizer [63] with a learning rate of 0.001 and a mini-batch size of 32. Adam optimizer tries to minimize mean absolute error (MAE) instead of mean squared error (MSE) because MAE is more robust in training the model with noisy training data. Adam optimizer [63] with a learning rate of 0.001, and a mini-batch size of 32 were applied to train the DNN and CNN model with regards to minimizing MAE. The formulated optimization models were implemented in Python 3. Then, they were solved with the Gurobi optimizer version 9.0. Ansarifar on Page 10: “The results of using the RIO to estimate uncertainty in the prediction of the GDU into the future (next two years) for sites 0 and 1 were visualized in Figure 9 and 10, respectively. The proposed time-series model was trained on data up to the end of 2019, and the predictions started on the first of 2020, and it then predicted the GDU 730 days into the future. To model uncertainty of weather for the next planting and harvesting calendar (2020 and 2021), the RIO approach was used to create 25 weather scenarios. The shadow areas represent the confidence interval of weather prediction, which indicates the range of variability across 25 weather scenarios. We used 25 weather scenarios to formulate the stochastic optimization model under weather uncertainty on the given calendar day of 2020.”), wherein the predictions comprise at least one or more coefficient values of one or more terms of the optimizer (see at least Ansarifar: Figs. 11-16 & Pages 8-10. Ansarifar teaches comparing the performance of the proposed time-series structure with the state-of-the-art in terms of three criteria: RMSE, which indicates the difference between predicted and observed weather, relative RMSE (RRMSE), which represents the normalized difference between predicted and observed weather, and coefficient of determination (R2), which computes the proportion of the variance in the weather that is explained by independent variables. See also Figs. 11-13 and Table 2: The aim of the optimization model in case 2 (sites have no capacities) is to schedule the seed population’s planting time at the lowest capacity required for both sites as well as the fewest number of weeks. Solving the optimization model (7)-(11) for both sites suggested that the lowest capacity required for sites 0 and 1 are 10,658 and 7,875. Then, the optimization model (1)-(5) was solved for various harvesting periods with determined capacities. For this purpose, we limited the model to determine the seeds’ planting times so that their harvests happened in the limited harvesting periods. The best harvesting periods in terms of minimizing Equation (6) are reported in Figure 14 for sites 0 and 1.);
- executing, using the at least one processing device, the optimizer (see at least Ansarifar: Pages 7-8 under “Experiment Settings”. Ansarifar teaches that to tune the parameter, we used a time-wise five-fold cross-validation, as shown in Figure 6. Each fold corresponds to particular six months as test data for prediction, and the rest of the data from 2009 up to test data corresponds to the training set. We applied Adam optimizer [63] with a learning rate of 0.001 and a mini-batch size of 32. Adam optimizer tries to minimize mean absolute error (MAE) instead of mean squared error (MSE) because MAE is more robust in training the model with noisy training data. Adam optimizer [63] with a learning rate of 0.001, and a mini-batch size of 32 were applied to train the DNN and CNN model with regards to minimizing MAE. The formulated optimization models were implemented in Python 3. Then, they were solved with the Gurobi optimizer version 9.0.) to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon for the underlying system (see at least Ansarifar: Fig. 6 & Figs. 11-16 & Pages 8-10 under Prediction accuracy comparison with other ML models. Ansarifar teaches that the proposed structure’s performance in predicting weather into the future and the optimization models’ performance in scheduling the planting time of seed population with more consistent harvest quantity among all weeks. See also Ansarifar at Pages 6-7 under Optimization Model noting that the scheduling problem as the optimization model using the predicted GDU as the heuristic measurement for harvesting. For case 1, the optimization model tried to schedule the planting of seed population at a minimum number of harvest weeks so that there is consistent harvest quantity among all weeks, and the capacity constraints are met. While at case 2, the optimization model determines the optimal scheduling of seed population’s planting time at the lowest possible capacity required. Two sites do not have interaction with each other, and we can optimize them separately. Moreover, we developed one optimization problem for case 1 (when sites have storage capacity) and one optimization for case 2 (when sites do not have storage capacity). In the following, the variables and parameters used in the model are described. Set of days in planting horizon, t ∈ T = {1,...,T}; Set of weeks in planting horizon, t ∈ W = {1,...,W}.) See also Figs. 11-13 and Table 2: The aim of the optimization model in case 2 (sites have no capacities) is to schedule the seed population’s planting time at the lowest capacity required for both sites as well as the fewest number of weeks. Solving the optimization model (7)-(11) for both sites suggested that the lowest capacity required for sites 0 and 1 are 10,658 and 7,875. Then, the optimization model (1)-(5) was solved for various harvesting periods with determined capacities. For this purpose, we limited the model to determine the seeds’ planting times so that their harvests happened in the limited harvesting periods. The best harvesting periods in terms of minimizing Equation (6) are reported in Figure 14 for sites 0 and 1.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff method for machine learning based production optimizers with the aforementioned teachings of: providing, using the at least one processing device, the predictions to an optimizer, wherein the predictions comprise at least one or more coefficient values of one or more terms of the optimizer & executing, using the at least one processing device, the optimizer to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon for the underlying system, and in view of Ansarifar, et. al., whereby the system of Ansarifar uses optimizers in order to develop a new framework that consists of a weather time series model and an optimization model to schedule the planting time. A deep recurrent neural network was designed to predict the weather into the future, and a Gaussian process model on top of the time-series model was developed to model the uncertainty of forecasted weather. The proposed optimization models also scheduled the seed population’s planting time at the fewest number of weeks with a more consistent weekly harvest quantity. Using the proposed optimization models can decrease the required capacity by 69% at site 0 and up to 51% at site 1 compared to the original planting time (see at least Ansarifar: ¶ [abstract].)
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Ansarifar, et. al, the results of the combination were predictable.
Moreover, regarding Independent Claim 1, Ruff / Ansarifar method for machine learning-based production optimizers does not explicitly disclose, but Kapadi in the analogous art for the method machine learning-based production optimizers does disclose the following:
- generating one or more schedules to operate one or more processing facilities (see at least Kapadi: Fig. 2 & ¶ [0087] & ¶ [0109].), based, at least in part, on the optimization results (see at least Kapadi: ¶ [0052-0053] & ¶ [0080-0082] & Fig. 2. Kapadi teaches that the sugar recovery and sugarcane yield models 20 and 22 as described above are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns. This optimization framework is described below. The harvest year is given by HDmin, which is the numerical equivalent of the start date of a harvest year, and by HDmax; which is the numerical equivalent of the end date of the harvest year. As indicated, these four parameters are given by a number equivalent to a date. The reference date used to calculate the number equivalent is arbitrarily chosen as 1 Jan. 2000, which is represented as 36526 by its number equivalent.)
- causing the one or more processing facilities to operate (see at least Kapadi: Fig. 2 & ¶ [0087] & ¶ [0109]. Kapadi notes that the plant data capacity is changed at 38 to reflect a change, such as an increase; in processing plant capacity. Then, the analysis performed at 24, 30, 40, 42, 44, and 38 is performed iteratively until a satisfactory return on investment is determined at 38, at which point an optimum processing capacity 46 is determined and provided as another output. This optimum processing capacity 46 may indicate, for example, that the processing plant that processes the sugarcane should be expanded in order to increase its capacity and thereby maximize return on investment. See also Kapadi at ¶ [0087]: If sugarcane planting is not sufficient to run the processing plant during the special harvest season, the processing plant is run only during the main harvest season. Hence, the main harvest season should always start, but starting of the special harvest season is optional. See also Kapadi at ¶ [0100]: The harvest and planting scheduling framework explained above focuses on harvest schedule generation when data about planted loads is known. However, the planting scheduling can be generated (meaning values for APv,pd are not known but need to be determined) by making simple modifications in the given formulation. The harvest and planting scheduling model can be run multiple times to find optimal capacity or for analysis of different what-if scenarios.) based, at least in part, on the one or more schedules (see at least Kapadi: Fig. 2 & Fig. 4 & ¶ [0100-0102] & ¶ [0108]. Kapadi notes that the sugar recovery model produced by the sugar recovery modeling procedure 20 and the sugarcane yield model produced by the sugarcane yield modeling procedure 22 are used by the optimizing procedure 24 of FIG. 2 in order to optimize a planting schedule 26, to optimize a harvesting schedule 28, and to maximize net farm returns 30. As inputs, the optimizing procedure 24 receives planting and harvesting practices data 32, observed and forecasted data 34, and ROI model data 36. The planting and harvesting practices data 32 includes data indicating the planting year, the harvest year including the main harvest season and the special harvest season, maintenance and festival days, number of varieties, sugarcane ages, etc. that serve to execute the sugar recovery and sugarcane yield models and to optimize the planting schedule 26 and the harvesting schedule 28 and to maximize the net farm returns 30. The optimum planting schedule 26 is produced using some unknown values for APv,pd and some known values for such parameters as main and special seasons operating days within a framework such as that provided by one of more of equations (65)-(86).)
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar method for machine learning based production optimizers with the aforementioned teachings of: generating one or more schedules to operate one or more processing facilities based, at least in part, on the optimization results and causing the one or more processing facilities to operate based, at least in part on the one or more schedules, and in further view of Kapadi, in order for the sugar recovery and sugarcane yield models 20 and 22 as described are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns (see at least Kapadi: ¶ [0080]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Kapadi, the results of the combination were predictable.
Regarding Independent Claims 15 and 29, Ruff apparatus / non-transitory computer readable medium for a machine learning-based production optimizers teaches the following:
- at least one processing device configured to (see at least Ruff: ¶ [0114-0116] & Fig. 1 & Fig. 4.);
- storing computer readable program code that, when executed by one or more processors, causes the one or more processors to (see at least Ruff: ¶ [0114-0116] & Fig. 1 & Fig. 4.);
- obtain (see at least Ruff: ¶ [0083] & ¶ [0091] & ¶ [0198]. Ruff teaches that the system 130 may obtain or ingest data under user 102 control, on a mass basis from a large number of growers who have contributed data to a shared database system. This form of obtaining data may be termed “manual data ingest” as one or more user-controlled computer operations are requested or triggered to obtain data for use by the system 130. See also Ruff at ¶ [0083]: The nitrogen map may include one or more user input features, such as dials or slider bars, to dynamically change the nitrogen planting and practices programs so that a user may optimize his nitrogen map, such as to obtain a preferred amount of surplus to shortfall. The user may then use his optimized nitrogen map and the related nitrogen planting and practices programs to implement one or more scripts, including variable rate (VR) fertility scripts. Similar instructions to the nitrogen instructions 210 could be used for application of other nutrients (such as phosphorus and potassium), application of pesticide, and irrigation programs. See also Ruff at ¶ [0198]: The system 130 is programmed to present summaries, tips, or further recommendations generated from analyzing the data obtained from the multitude of prescribed experiments across grower fields.) data from one or more data sources (see at least Ruff: ¶ [0107-0108] & ¶ [0201]. Ruff teaches that the agricultural intelligence computer system 130 is configured or programmed to implement agronomic data preprocessing of field data received from one or more data sources. The field data received from one or more data sources may be preprocessed for the purpose of removing noise, distorting effects, and confounding factors within the agronomic data including measured outliers that could adversely affect received field data values. The historical data can be obtained from internal trials and experiments or from external data sources. The number of fields can have common values in certain characteristics, such as the crop hybrid grown in a field, the location of a field, or the yield lift management practice for a field.), wherein the data is associated with or affects an underlying system to be optimized (see at least Ruff: ¶ [0105-0107] & ¶ [0109] & ¶ [0201]. Ruff teaches that the agricultural intelligence computer system 130 is programmed or configured to create an agronomic model. In this context, an agronomic model is a data structure in memory of the agricultural intelligence computer system 130 that comprises field data 106, such as identification data and harvest data for one or more fields. The agronomic model may also comprise calculated agronomic properties which describe either conditions which may affect the growth of one or more crops on a field, or properties of the one or more crops, or both. seeds and planting instructions 208 are programmed to provide tools for seed selection, hybrid placement, and script creation, including variable rate (VR) script creation, based upon scientific models and empirical data. This enables growers to maximize yield or return on investment through optimized seed purchase, placement and population. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method.);
- generate (see at least Ruff: ¶ [0194-0197] & ¶ [0201] & ¶ [0208-0211].) predictions based on the obtained data, wherein the predictions represent estimated values comprising numeric values of one or more time-varying parameters (see at least Ruff: ¶ [0105-0106] & ¶ [0212-0123]. Ruff teaches that the system 130 is programmed to further analyze the data, to adjust the predictions or the plans for the prescribed experiments, or to glean specific insight that can be used in designing future experiments. Such analysis can be performed periodically, at the end of a season or a year, or upon request by a grower. When a prescribed experiment was not properly carried out, the predicted result might not be obtained, and the system 130 can be programmed to adjust the prediction based on how the plan for the prescribed experiment was followed. For example, the system 130 can be configured to consider that the actual seeding rate was only 80% of the prescribed seeding rate overall, due to erroneous calibration of the agricultural implement, the skipping of certain planting steps, or other reasons, in determining the predicted crop yield might be only 80% of or otherwise less than the predicted or recommended crop yield. The system 130 can also be programmed to generate a series of remedial steps in order to realize the original prediction. For example, when the actual seeding rate was only 80% of the prescribed seeding rate overall, the system 130 can be configured to compensate for it by prescribing a seeding rate that was 20% or otherwise higher than originally prescribed for the rest of the experiment. See also Ruff at ¶ [0105]: The agronomic factors may also be used to estimate one or more crop related results, such as agronomic yield. The agronomic yield of a crop is an estimate of quantity of the crop that is produced, or in some examples the revenue or profit obtained from the produced crop. See also Ruff at ¶ [0211]: The group of fields may be selected from those fields that share values with the grower's field in certain characteristics, such as the crop hybrid grown in a field, the predicted yield lift for a change in management practice for a field, or the location of a field. See also Ruff at ¶ [0242]: The agricultural intelligence computing system may additionally alter the predicted results of the trial based on identified modifications to the testing locations. For example, the agricultural intelligence computing system may predict an increase in yield of 30 bushels/acre for an application of 40 lbs/acre of nitrogen. If the agricultural intelligence computing system detects that only 30 lbs/acre of nitrogen has been applied to a field, the agricultural intelligence computing system may lower the predicted increase in yield of 30 bushels/acre.).) associated with the underlying system at a plurality of time periods in an optimization horizon (see at least Ruff: ¶ [0185] & ¶ [0191-0194] & ¶ [0211]. Ruff teaches that the objective is to enable proper implementation of the prescribed experiments in order to achieve the predicted results. For the variables involved in the scheme, the system 130 is programmed to compare the actual value, such as the volume of seeds actually planted at a specific location within a particular period of time, such as one hour, and the prescribed value. See also Ruff at ¶ [0194]: The system 130 is programmed to further analyze the data, to adjust the predictions or the plans for the prescribed experiments, or to glean specific insight that can be used in designing future experiments. Such analysis can be performed periodically, at the end of a season or a year, or upon request by a grower. See also Ruff at ¶ [0211]: In step 1702, the system 130 is programmed to receive crop seeding rate data and corresponding crop yield data over a period of time regarding a group of fields associated with a plurality of grower devices. Such data is used to establish benchmarks for determining product responsiveness to a seeding rate increase for a grower's field. The group of fields may be selected from those fields that share values with the grower's field in certain characteristics, such as the crop hybrid grown in a field, the predicted yield lift for a change in management practice for a field, or the location of a field.)
Regarding Independent Claims 15 and 29, Ruff apparatus / non-transitory computer readable medium for machine learning-based production optimizers does not explicitly disclose, but Ansarifar, et. al. in the analogous art for the apparatus / non-transitory computer readable medium machine learning-based production optimizers does disclose the following:
- provide the predictions to an optimizer (see at least Ansarifar: Figs. 7-8 & Pages 7-8 under “Experiment Settings”. Ansarifar teaches that to tune the parameter, we used a time-wise five-fold cross-validation, as shown in Figure 6. Each fold corresponds to particular six months as test data for prediction, and the rest of the data from 2009 up to test data corresponds to the training set. We applied Adam optimizer [63] with a learning rate of 0.001 and a mini-batch size of 32. Adam optimizer tries to minimize mean absolute error (MAE) instead of mean squared error (MSE) because MAE is more robust in training the model with noisy training data. Adam optimizer [63] with a learning rate of 0.001, and a mini-batch size of 32 were applied to train the DNN and CNN model with regards to minimizing MAE. The formulated optimization models were implemented in Python 3. Then, they were solved with the Gurobi optimizer version 9.0. Ansarifar on Page 10: “The results of using the RIO to estimate uncertainty in the prediction of the GDU into the future (next two years) for sites 0 and 1 were visualized in Figure 9 and 10, respectively. The proposed time-series model was trained on data up to the end of 2019, and the predictions started on the first of 2020, and it then predicted the GDU 730 days into the future. To model uncertainty of weather for the next planting and harvesting calendar (2020 and 2021), the RIO approach was used to create 25 weather scenarios. The shadow areas represent the confidence interval of weather prediction, which indicates the range of variability across 25 weather scenarios. We used 25 weather scenarios to formulate the stochastic optimization model under weather uncertainty on the given calendar day of 2020.”), wherein the predictions comprise at least one or more coefficient values of one or more terms of the optimizer (see at least Ansarifar: Figs. 11-16 & Pages 8-10. Ansarifar teaches comparing the performance of the proposed time-series structure with the state-of-the-art in terms of three criteria: RMSE, which indicates the difference between predicted and observed weather, relative RMSE (RRMSE), which represents the normalized difference between predicted and observed weather, and coefficient of determination (R2), which computes the proportion of the variance in the weather that is explained by independent variables. See also Figs. 11-13 and Table 2: The aim of the optimization model in case 2 (sites have no capacities) is to schedule the seed population’s planting time at the lowest capacity required for both sites as well as the fewest number of weeks. Solving the optimization model (7)-(11) for both sites suggested that the lowest capacity required for sites 0 and 1 are 10,658 and 7,875. Then, the optimization model (1)-(5) was solved for various harvesting periods with determined capacities. For this purpose, we limited the model to determine the seeds’ planting times so that their harvests happened in the limited harvesting periods. The best harvesting periods in terms of minimizing Equation (6) are reported in Figure 14 for sites 0 and 1.);
- execute the optimizer (see at least Ansarifar: Pages 7-8 under “Experiment Settings”. Ansarifar teaches that to tune the parameter, we used a time-wise five-fold cross-validation, as shown in Figure 6. Each fold corresponds to particular six months as test data for prediction, and the rest of the data from 2009 up to test data corresponds to the training set. We applied Adam optimizer [63] with a learning rate of 0.001 and a mini-batch size of 32. Adam optimizer tries to minimize mean absolute error (MAE) instead of mean squared error (MSE) because MAE is more robust in training the model with noisy training data. Adam optimizer [63] with a learning rate of 0.001, and a mini-batch size of 32 were applied to train the DNN and CNN model with regards to minimizing MAE. The formulated optimization models were implemented in Python 3. Then, they were solved with the Gurobi optimizer version 9.0.) to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon for the underlying system (see at least Ansarifar: Fig. 6 & Figs. 11-16 & Pages 8-10 under Prediction accuracy comparison with other ML models. Ansarifar teaches that the proposed structure’s performance in predicting weather into the future and the optimization models’ performance in scheduling the planting time of seed population with more consistent harvest quantity among all weeks. See also Ansarifar at Pages 6-7 under Optimization Model noting that the scheduling problem as the optimization model using the predicted GDU as the heuristic measurement for harvesting. For case 1, the optimization model tried to schedule the planting of seed population at a minimum number of harvest weeks so that there is consistent harvest quantity among all weeks, and the capacity constraints are met. While at case 2, the optimization model determines the optimal scheduling of seed population’s planting time at the lowest possible capacity required. Two sites do not have interaction with each other, and we can optimize them separately. Moreover, we developed one optimization problem for case 1 (when sites have storage capacity) and one optimization for case 2 (when sites do not have storage capacity). In the following, the variables and parameters used in the model are described. Set of days in planting horizon, t ∈ T = {1,...,T}; Set of weeks in planting horizon, t ∈ W = {1,...,W}.) See also Figs. 11-13 and Table 2: The aim of the optimization model in case 2 (sites have no capacities) is to schedule the seed population’s planting time at the lowest capacity required for both sites as well as the fewest number of weeks. Solving the optimization model (7)-(11) for both sites suggested that the lowest capacity required for sites 0 and 1 are 10,658 and 7,875. Then, the optimization model (1)-(5) was solved for various harvesting periods with determined capacities. For this purpose, we limited the model to determine the seeds’ planting times so that their harvests happened in the limited harvesting periods. The best harvesting periods in terms of minimizing Equation (6) are reported in Figure 14 for sites 0 and 1.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: provide the predictions to an optimizer, wherein the predictions comprise at least one or more coefficient values of one or more terms of the optimizer & execute the optimizer to generate optimization results based on the predictions, the optimization results respectively associated with the plurality of time periods in the optimization horizon for the underlying system, and in view of Ansarifar, et. al., whereby the system of Ansarifar uses optimizers in order to develop a new framework that consists of a weather time series model and an optimization model to schedule the planting time. A deep recurrent neural network was designed to predict the weather into the future, and a Gaussian process model on top of the time-series model was developed to model the uncertainty of forecasted weather. The proposed optimization models also scheduled the seed population’s planting time at the fewest number of weeks with a more consistent weekly harvest quantity. Using the proposed optimization models can decrease the required capacity by 69% at site 0 and up to 51% at site 1 compared to the original planting time (see at least Ansarifar: ¶ [abstract].)
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Ansarifar, et. al, the results of the combination were predictable.
Moreover, regarding Independent Claims 15 and 29, Ruff / Ansarifar apparatus / non-transitory computer readable medium for machine learning-based production optimizers does not explicitly disclose, but Kapadi in the analogous art for the apparatus machine learning-based production optimizers does disclose the following:
- generate one or more schedules to operate one or more processing facilities (see at least Kapadi: Fig. 2 & ¶ [0087] & ¶ [0109].), based, at least in part, on the optimization results (see at least Kapadi: ¶ [0052-0053] & ¶ [0080-0082] & Fig. 2. Kapadi teaches that the sugar recovery and sugarcane yield models 20 and 22 as described above are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns. This optimization framework is described below. The harvest year is given by HDmin, which is the numerical equivalent of the start date of a harvest year, and by HDmax; which is the numerical equivalent of the end date of the harvest year. As indicated, these four parameters are given by a number equivalent to a date. The reference date used to calculate the number equivalent is arbitrarily chosen as 1 Jan. 2000, which is represented as 36526 by its number equivalent.);
- cause the one or more processing facilities to operate (see at least Kapadi: Fig. 2 & ¶ [0087] & ¶ [0109]. Kapadi notes that the plant data capacity is changed at 38 to reflect a change, such as an increase; in processing plant capacity. Then, the analysis performed at 24, 30, 40, 42, 44, and 38 is performed iteratively until a satisfactory return on investment is determined at 38, at which point an optimum processing capacity 46 is determined and provided as another output. This optimum processing capacity 46 may indicate, for example, that the processing plant that processes the sugarcane should be expanded in order to increase its capacity and thereby maximize return on investment. See also Kapadi at ¶ [0087]: If sugarcane planting is not sufficient to run the processing plant during the special harvest season, the processing plant is run only during the main harvest season. Hence, the main harvest season should always start, but starting of the special harvest season is optional. See also Kapadi at ¶ [0100]: The harvest and planting scheduling framework explained above focuses on harvest schedule generation when data about planted loads is known. However, the planting scheduling can be generated (meaning values for APv,pd are not known but need to be determined) by making simple modifications in the given formulation. The harvest and planting scheduling model can be run multiple times to find optimal capacity or for analysis of different what-if scenarios.) based, at least in part, on the one or more schedules (see at least Kapadi: Fig. 2 & Fig. 4 & ¶ [0100-0102] & ¶ [0108]. Kapadi notes that the sugar recovery model produced by the sugar recovery modeling procedure 20 and the sugarcane yield model produced by the sugarcane yield modeling procedure 22 are used by the optimizing procedure 24 of FIG. 2 in order to optimize a planting schedule 26, to optimize a harvesting schedule 28, and to maximize net farm returns 30. As inputs, the optimizing procedure 24 receives planting and harvesting practices data 32, observed and forecasted data 34, and ROI model data 36. The planting and harvesting practices data 32 includes data indicating the planting year, the harvest year including the main harvest season and the special harvest season, maintenance and festival days, number of varieties, sugarcane ages, etc. that serve to execute the sugar recovery and sugarcane yield models and to optimize the planting schedule 26 and the harvesting schedule 28 and to maximize the net farm returns 30. The optimum planting schedule 26 is produced using some unknown values for APv,pd and some known values for such parameters as main and special seasons operating days within a framework such as that provided by one of more of equations (65)-(86).)
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: generate one or more schedules to operate one or more processing facilities based, at least in part, on the optimization results and cause the one or more processing facilities to operate based, at least in part on the one or more schedules, and in further view of Kapadi, in order for the sugar recovery and sugarcane yield models 20 and 22 as described are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns (see at least Kapadi: ¶ [0080]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Kapadi, the results of the combination were predictable.
Regarding Dependent Claims 3, 17 and 31, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Independent Claims 1, 15 and 29 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- wherein the optimizer (see at least Ruff: ¶ [0104-0106] noting the trained model and Ruff ¶ [0201-0204] system 130 as the optimizer.) is configured to perform production schedule optimization (see at least Ruff: ¶ [0186] & ¶ [0193] & Fig. 14-16. Ruff teaches that the system 130 is programmed to prescribe experiments to grower fields and the design or selection of experiments can be carried out automatically according to a predetermined schedule, such as at the beginning of every year or every growing season. The prescribing of experiments can also be performed automatically. The system 130 can be configured to generate the prescription, plan, or scheme for an experiment that is to be understood by a human, a machine, or a combination of both. See also Ruff at ¶ [0051] showing the Actual Production History (APH). See also Ruff at ¶ [0193]: The system 130 can be programmed to validate the execution of each prescribed experiment according to a predetermined schedule, such as every month, or as soon as error signals or application data are received. Some optimization results depicted at Ruff of Figs. 14-15 and Fig. 16 elements 1612, 1614, 1616 and associated text.).
Regarding Dependent Claims 4, 18 and 32, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Claims 1, 3, 15, 17, 29 and 31 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- the underlying system comprises an agricultural system in which crops grow in multiple growing areas (see at least Ruff: Fig. 8 & ¶ [0130] & ¶ [0218-0220]. Ruff teaches an agricultural system in which crops grow in multiple zones shown in Fig. 8. See also Ruff at ¶ [0130]: The trial may require one or more testing locations to be placed in an area of the field with conditions differing from the rest of the field and/or areas of the field spanning different types of conditions. The trial may require one or more different management practices to be undertaken in one or more testing locations. See also Ruff at ¶ [0135]: The agricultural intelligence computing system may identify locations on the field for implementing a test location based on areas in the field capable of performing the trial, efficiency of performing the trial in each location, applicability of the trial to other locations, and/or benefit to the field of performing the trial. See also Ruff at ¶ [0161]: An agricultural intelligence computing system may determine the responsiveness of different areas for a particular field based on prior practices, prior yield data, and other field data from one or many fields. The agricultural intelligence computing system may then determine the effectiveness of applying the product to the responsive portions and the non-responsive portions of the field. See also Ruff at ¶ [0218]: Management zones are identified based on both responsiveness and total yield. The agricultural intelligence computing system may determine the responsiveness of areas in a field to applications of products and/or different management practices based on prior yield data, soil data, imagery, other crop data, and management practices. See also Ruff at ¶ [0225]: For example, the agricultural intelligence computing system may prioritize areas of the field that have had historically lower yields, thereby reducing any possible negative impacts on the yield of the field.);
- the optimization results (see at least Ruff: ¶ [0200-0208] & ¶ [0220]. Ruff notes the use of the field targeting and based on the inputs that have been fed into the system at the prior limitations. Some optimization results are depicted at Figs. 14-15 and Fig. 16 elements 1612, 1614, 1616 and associated text.) comprises a schedule identifying resources that are scheduled to perform harvesting of the crops in the multiple growing areas or in lots within the multiple growing areas (see at least Ruff: ¶ [0146-0147] & ¶ [0186-0190] & Fig. 8. Ruff notes that the farm equipment use history may include identification of the tilling, planting, application, and harvesting equipment. The field operator data may identify one or more people, operations, or service providers who perform activities on the field. See also identification of processing facility, and/or identification of one or more people processing and/or collecting the soil, additional soil chemistry data, bulk density of the soil, and/or buffer capacity. The design or selection of experiments can be carried out automatically according to a predetermined schedule, such as at the beginning of every year or every growing season. The plan for the experiment can include a variety of details, such as the type of seeds, the destination of the seeds within the fields, the volume of seeds to plant each day, or the time to plant the seeds each day. See also Ruff noting growing areas or growing zones of harvesting of crops shown in Figs. 8-10.) and when the resources scheduled to perform harvesting of the crops in the multiple growing areas or in the lots within the multiple growing areas (see at least Ruff: ¶ [0146-0147] & ¶ [0186-0190] & ¶ [0225]. Ruff notes that the system 130 can be configured to generate the prescription, plan, or scheme for an experiment that is to be understood by a human, a machine, or a combination of both. For example, one experiment may be to plant certain seeds at certain rates on a certain grower's fields. The plan for the experiment can include a variety of details, such as the type of seeds, the destination of the seeds within the fields, the volume of seeds to plant each day, or the time to plant the seeds each day. See also Ruff at ¶ [0225]: Instead of transmitting the entire scheme for an experiment to a smart device, whether it is an agricultural implement or a person digital assistant, the system 130 is programmed to transmit the scheme incrementally and timely. For example, when the scheme involves the performance of daily tasks, the system 130 can be configured to send a portion of the scheme corresponding to each day's work every day. See also Ruff noting growing areas or growing zones of harvesting of crops shown in Figs. 8-10.)
Moreover, regarding Dependent Claims 4, 18 and 32, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose in Kapadi in the analogous art for method / apparatus / non-transitory computer readable medium for machine learning based production optimizers the following:
- the one or more time-varying parameters relate to one or more products recoverable in the crops (see at least Kapadi: ¶ [abstract] & ¶ [0007] & Fig. 2. Kapadi notes that sugarcane yield and sugar recovery within a sugar mill region vary with a combination of deterministic parameters (e.g., variety, crop class or ratoon type, age, harvest date indicating season) and stochastic parameters (e.g., weather conditions, soil type, farming practices, irrigation facilities). See also Kapadi at ¶ [abstract]: A harvesting and/or planting schedule is generated based on a product recovery model and a crop yield model. The product recovery model models recovery of a product, such as sugar, from a crop, such as sugarcane. The crop yield model models yield of the crop from land. First, second, third, and fourth data are used to generate the harvesting and/or planting schedule. The first input data is pertinent to predicting the recovery of the product by use of the product recovery model. See also Kapadi at ¶ [0029]: The combined model represented by equations (4) and (5) can be fitted to the production and harvest training data of an industry in order to estimate the parameters av, bv, cv, dv, and ev. See also Kapadi at Fig. 2 noting “recovery estimation models - 20” and “yield estimation models – 22”.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: the one or more time-varying parameters relate to one or more products recoverable in the crops, and in further view of Kapadi, in order for the sugar recovery and sugarcane yield models 20 and 22 as described are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns (see at least Kapadi: ¶ [0080]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Kapadi, the results of the combination were predictable.
Regarding Dependent Claims 6, 20 and 34, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Independent Claims 1, 15 and 29 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- wherein the optimizer (see at least Ruff: ¶ [0104-0106] noting the trained model and Ruff ¶ [0201-0204] system 130 as the optimizer.) is configured to perform process optimization (see at least Ruff: ¶ [0147] & ¶ [0186-0190] & Figs. 14-16. Ruff notes identification of processing facility, and/or identification of one or more people processing and/or collecting the soil, additional soil chemistry data, bulk density of the soil, and/or buffer capacity. See also Ruff at ¶ [0186]: The design or selection of experiments can be carried out automatically according to a predetermined schedule, such as at the beginning of every year or every growing season. The prescribing of experiments can also be performed automatically. The system 130 can be configured to generate the prescription, plan, or scheme for an experiment that is to be understood by a human, a machine, or a combination of both. For example, one experiment may be to plant certain seeds at certain rates on a certain grower's fields. See also Ruff at ¶ [0190]: The system 130 is programmed to receive data from the same agricultural implements to which the experiment schemes or plans were transmitted, or from the same field manager computing device, including mobile devices, registered under the growers. The agricultural implements can be equipped with sensors that can capture many types of data. In addition to data related to the variables involved in the experiment, such as the volume of seeds actually planted, the time of actual planting, the actual moving or rotational speed of the agricultural implement, the route actually taken by the agricultural implement, or the crop yield actually achieved. The data can be transmitted by an agricultural implement or a personal computing device to the system 130 once the data becomes available, upon request by the system 130, or according to a predetermined schedule.).
Regarding Dependent Claims 7, 21 and 35, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Claims 1, 6, 15, 20, 29 and 34 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- the optimization results (see at least Ruff: ¶ [0200-0208] & ¶ [0220]. Ruff notes the use of the field targeting and based on the inputs that have been fed into the system at the prior limitations. Some optimization results are depicted at Figs. 14-15 and Fig. 16 elements 1612, 1614, 1616 and associated text.) comprise one or more settings for equipment in the processing facility (see at least Ruff: ¶ [0090] & ¶ [0146-0147] & ¶ [0154]. Ruff teaches that application controller 114 may also be programmed or configured to control an operating parameter of an agricultural vehicle or implement. For example, an application controller may be programmed or configured to control an operating parameter of a vehicle, such as a tractor, planting equipment, tillage equipment, fertilizer or insecticide equipment, harvester equipment, or other farm implements such as a water valve. See also Ruff at ¶ [0146]: Farm equipment use history may include identification of the tilling, planting, application, and harvesting equipment. The field operator data may identify one or more people, operations, or service providers who perform activities on the field. See also Ruff at ¶ [0147] regarding “identification of processing facility.” See also Ruff at ¶ [0154]: Tiling data may include presence of tiling, tiling system types, tiling system maps, tiling system flow conductances, and/or flow rates or fluid levels in tile lines.).
Moreover, regarding Dependent Claims 7, 21 and 35, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose in Kapadi in the analogous art for method / apparatus / non-transitory computer readable medium for machine learning based production optimizers the following:
- the underlying system comprises a processing facility configured to process harvested crops (see at least Kapadi: ¶ [abstract] & ¶ [0087] & ¶ [0107]. Kapadi notes that the third input data relates to capacity of a crop processing plant to process the crop to produce the product. See also Kapadi at ¶ [0087]: If sugarcane planting is not sufficient to run the processing plant during the special harvest season, the processing plant is run only during the main harvest season. Hence, the main harvest season should always start, but starting of the special harvest season is optional. See also Kapadi at ¶ [0107]: The return on investment models 42 model the investments that affect processing plant capacity such as cost of facilities associated with the milling and harvesting of sugarcane, expected prices of sugar, molasses, and/or bagasse, transportation costs, storage costs, and other costs that affect the return on the investment associated with the production of sugar, molasses, and/or bagasse. See also Kapadi at ¶ [0112]: The output device(s) is capable of outputting the planting schedule 26, the harvesting schedule 28, and the net farm returns 30. The output device(s) 88 is also capable of outputting the processing plant capacity 46.);
- the one or more time-varying parameters relate to one or more products recoverable in the crops (see at least Kapadi: ¶ [abstract] & ¶ [0007] & Fig. 2. Kapadi notes that sugarcane yield and sugar recovery within a sugar mill region vary with a combination of deterministic parameters (e.g., variety, crop class or ratoon type, age, harvest date indicating season) and stochastic parameters (e.g., weather conditions, soil type, farming practices, irrigation facilities). See also Kapadi at ¶ [abstract]: A harvesting and/or planting schedule is generated based on a product recovery model and a crop yield model. The product recovery model models recovery of a product, such as sugar, from a crop, such as sugarcane. The crop yield model models yield of the crop from land. First, second, third, and fourth data are used to generate the harvesting and/or planting schedule. The first input data is pertinent to predicting the recovery of the product by use of the product recovery model. See also Kapadi at ¶ [0029]: The combined model represented by equations (4) and (5) can be fitted to the production and harvest training data of an industry in order to estimate the parameters av, bv, cv, dv, and ev. See also Kapadi at Fig. 2 noting “recovery estimation models - 20” and “yield estimation models – 22”.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: the underlying system comprises a processing facility configured to process harvested crops & the one or more time-varying parameters relate to one or more products recoverable in the crops, and in further view of Kapadi, in order for the sugar recovery and sugarcane yield models 20 and 22 as described are used by an optimizing procedure 24 of FIG. 2 according to an optimization framework in order to optimize the planting and harvesting schedules to maximize net farm returns (see at least Kapadi: ¶ [0080]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Kapadi, the results of the combination were predictable.
Regarding Dependent Claims 8, 22 and 36, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Independent Claims 1, 15 and 29 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- wherein the predictions are generated using a trained machine learning model (see at least Ruff: ¶ [0104] & ¶ [0203-0206] & ¶ [0220]. Ruff teaches a Process Overview-Agronomic Model Training. See also modeled yield variability data based on predictors to a model trained on historic yield variability data. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all-subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method. See also Ruff at ¶ [0203]: Other machine-learning methods known to someone skilled in the art for capturing various relationship between the seeding rate (in conjunction with other attributes) and the crop yield lift, such as neural networks or regression techniques can be used. See also Ruff at ¶ [0206]: The type of management practice can also be an input attribute for a machine learning method discussed above.).
Regarding Dependent Claims 9, 23 and 37, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Claims 1, 8, 15, 22, 29 and 36 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- wherein the trained machine learning model is trained (see at least Ruff: ¶ [0104] & ¶ [0203-0206] & ¶ [0220]. Ruff teaches a Process Overview-Agronomic Model Training. See also modeled yield variability data based on predictors to a model trained on historic yield variability data. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all-subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method. See also Ruff at ¶ [0203]: Other machine-learning methods known to someone skilled in the art for capturing various relationship between the seeding rate (in conjunction with other attributes) and the crop yield lift, such as neural networks or regression techniques can be used. See also Ruff at ¶ [0206]: The type of management practice can also be an input attribute for a machine learning method discussed above.) to generate the predictions when the obtained data includes imperfections (see at least Ruff: ¶ [0108] & ¶ [0196]. Ruff teaches that the field data received from one or more data sources may be preprocessed for the purpose of removing noise, distorting effects, and confounding factors within the agronomic data including measured outliers that could adversely affect received field data values. Embodiments of agronomic data preprocessing may include, but are not limited to, removing data values commonly associated with outlier data values. See also Ruff at [0196]: The system 130 can also be configured to detect patterns from the outcomes of similar experiments, which can help identify outliers and point to field-specific issues. The reasons behind the discrepancies between the predicted outcomes and the actual outcomes can be used for designing future experiments or generating predictions for future experiments. Examiner Note: Examiner interprets that the “imperfections” are outliers determined when performing predictions when the ML model is trained.).
Regarding Dependent Claims 12, 26 and 40, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Independent Claims 1, 15 and 29 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- further comprising: iteratively obtaining the data (see at least Ruff: ¶ [0187] & ¶ [0199] & (Dependent Claim 8 of Ruff). Ruff notes that in one scheme, a grower's field can be divided into locations, and the prescription can indicate that the first location is to be used for the targeted trial, the second location is to be used for the control trial, and this pattern is to repeat three times geographically (the second time on the 3rd and fourth locations, and the 3 time on the 5th and the sixth locations). Some or all of these steps 1302 through 1312 can be executed repeatedly, iteratively, or out of order. For example, data capturing and execution validation can take place periodically during a season.), generating the predictions (see at least Ruff: ¶ [0194-0197] & ¶ [0201] & ¶ [0208-0211]. Ruff teaches that the system 130 is programmed to further analyze the data, to adjust the predictions or the plans for the prescribed experiments, or to glean specific insight that can be used in designing future experiments. Such analysis can be performed periodically, at the end of a season or a year, or upon request by a grower. When a prescribed experiment was not properly carried out, the predicted result might not be obtained, and the system 130 can be programmed to adjust the prediction based on how the plan for the prescribed experiment was followed. For example, the system 130 can be configured to consider that the actual seeding rate was only 80% of the prescribed seeding rate overall, due to erroneous calibration of the agricultural implement, the skipping of certain planting steps, or other reasons, in determining the predicted crop yield might be only 80% of or otherwise less than the predicted or recommended crop yield. The system 130 can also be programmed to generate a series of remedial steps in order to realize the original prediction. For example, when the actual seeding rate was only 80% of the prescribed seeding rate overall, the system 130 can be configured to compensate for it by prescribing a seeding rate that was 20% or otherwise higher than originally prescribed for the rest of the experiment. See also Ruff at ¶ [0105]: The agronomic factors may also be used to estimate one or more crop related results, such as agronomic yield. The agronomic yield of a crop is an estimate of quantity of the crop that is produced, or in some examples the revenue or profit obtained from the produced crop. See also Ruff at ¶ [0211]: The group of fields may be selected from those fields that share values with the grower's field in certain characteristics, such as the crop hybrid grown in a field, the predicted yield lift for a change in management practice for a field, or the location of a field. See also Ruff at ¶ [0242]: The agricultural intelligence computing system may additionally alter the predicted results of the trial based on identified modifications to the testing locations. For example, the agricultural intelligence computing system may predict an increase in yield of 30 bushels/acre for an application of 40 lbs/acre of nitrogen. If the agricultural intelligence computing system detects that only 30 lbs/acre of nitrogen has been applied to a field, the agricultural intelligence computing system may lower the predicted increase in yield of 30 bushels/acre.), providing the predictions to the optimizer (see at least Ruff: ¶ [0104-0106] & ¶ [0109] & ¶ [0201-0204]. Ruff teaches that the agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method. For example, a genetic algorithm selection technique uses an adaptive heuristic search algorithm, based on evolutionary principles of natural selection and genetics, to determine and evaluate datasets within the preprocessed agronomic data. See also Ruff at ¶ [0201-0204] for system 130 as the optimizer and/or the trained model of Ruff at paragraphs ¶ [0104-0106].), and executing the optimizer (see at least Ruff: ¶ [0200-0208] & ¶ [0220]. Ruff notes the use of the field targeting and based on the inputs that have been fed into the system at the prior limitations. Some optimization results are depicted at Figs. 14-15 and Fig. 16 elements 1612, 1614, 1616 and associated text.).
- wherein at least one of the optimization results or the predictions from one iteration are provided as feedback for use during generation of the predictions during a subsequent iteration (see at least Ruff: ¶ [0108-0110] & (Dependent Claim 8 of Ruff). Ruff notes that the agronomic dataset evaluation logic is used as a feedback loop where agronomic datasets that do not meet configured quality thresholds are used during future data subset selection steps (block 310). The report can also outline possible experiments to apply to the grower's fields in the future and solicit feedback from the grower. See also (Dependent Claim 8 of Ruff): “Updating one or more predictions of a result of the trial on the agricultural field based, at least in part, on the one or more parameters of the application.”).
Regarding Dependent Claims 13, 27 and 41, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Claims 1, 12, 15, 26, 29 and 40 above, and Ruff further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- the predictions are generated using a trained machine learning model (see at least Ruff: ¶ [0104] & ¶ [0203-0206] & ¶ [0220]. Ruff teaches a Process Overview-Agronomic Model Training. See also modeled yield variability data based on predictors to a model trained on historic yield variability data. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all-subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method. See also Ruff at ¶ [0203]: Other machine-learning methods known to someone skilled in the art for capturing various relationship between the seeding rate (in conjunction with other attributes) and the crop yield lift, such as neural networks or regression techniques can be used. See also Ruff at ¶ [0206]: The type of management practice can also be an input attribute for a machine learning method discussed above.);
- the trained machine learning model (see at least Ruff: ¶ [0104] & ¶ [0203-0206] & ¶ [0220]. Ruff teaches a Process Overview-Agronomic Model Training. See also modeled yield variability data based on predictors to a model trained on historic yield variability data. See also Ruff at ¶ [0109]: The agricultural intelligence computer system 130 may implement data subset selection techniques including, but not limited to, a genetic algorithm method, an all-subset models method, a sequential search method, a stepwise regression method, a particle swarm optimization method, and an ant colony optimization method. See also Ruff at ¶ [0203]: Other machine-learning methods known to someone skilled in the art for capturing various relationship between the seeding rate (in conjunction with other attributes) and the crop yield lift, such as neural networks or regression techniques can be used. See also Ruff at ¶ [0206]: The type of management practice can also be an input attribute for a machine learning method discussed above.) is configured to use the feedback to compensate for prediction errors associated with the predictions (see at least Ruff: ¶ [0110] & ¶ [0193] & ¶ [0241]. Ruff notes that agronomic models may be compared and/or validated using one or more comparison techniques, such as, but not limited to, root mean square error with leave-one-out cross validation (RMSECV), mean absolute error, and mean percentage error. For example, RMSECV can cross validate agronomic models by comparing predicted agronomic property values created by the agronomic model against historical agronomic property values collected and analyzed. The agronomic dataset evaluation logic is used as a feedback loop where agronomic datasets that do not meet configured quality thresholds are used during future data subset selection steps (block 310). See also Ruff at ¶ [0193]: The system 130 can be programmed to validate the execution of each prescribed experiment according to a predetermined schedule, such as every month, or as soon as error signals or application data are received. See also Ruff at ¶ [0241]: The agricultural intelligence computing system may be programmed or configured to alter one or more trials in response to determining that a testing location does not comply with a trial. The agricultural intelligence computing system suggests alterations to one or more practices for other locations to offset errors in the testing location. If a control location was planted with a seeding rate that is ten percent higher than required by the trial, the agricultural intelligence computing system may modify the seeding rate for the other testing locations to be ten percent higher. See also (Dependent Claim 8 of Ruff): “Updating one or more predictions of a result of the trial on the agricultural field based, at least in part, on the one or more parameters of the application.”)
14. Claims 2, 16 and 30 are rejected under 35 U.S.C. 103 as being unpatentable over US PG Pub (US 2019/0057461 A1) hereinafter Ruff, et. al, in view of NPL Document: "Scheduling planting time through developing an optimization model and analysis of time series growing degree units." (2022), hereinafter Ansarifar, et. al., and in further view of US PG Pub (US 2009/0234695 A1) hereinafter Kapadi, et. al, and in further view of US PG Pub (US 2018/0181893 A1) hereinafter Basso.
Regarding Dependent Claims 2, 16 and 30, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does not explicitly disclose, but Basso teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose the following:
- wherein executing the optimizer comprises performing optimization (see at least Basso: & ¶ [0054] & ¶ [0078] & ¶ [0108]. Basso teaches that the optimized parameters can be determined on a whole-field basis or on a portion of the field (e.g., a higher target crop yield in one portion of the field and a relatively lower target crop yield in another portion of the field). If two or more field state properties are being optimized (e.g., two or more properties related to the same field quantity; two or more properties related to the different field quantities), a weighted and/or constrained optimization system can be used by the crop model 212 to determine the crop management plan 210. See also Basso at ¶ [0078]: When the actual yield 180 is within about 5%, 10%, 20%, 30%, or 50% of the model yield 340 (e.g., expressed on a whole-field basis or based on one or more sub-regions of the field), it can be inferred that the grower substantially followed the crop management plans 210, 220 determined during the growing season and satisfied any compliance parameters, which were included as optimization constraints for the crop models 212, 222 to determine the corresponding plans 210, 220. See also Basso at Figs. 1, 1A, 1B and 1C.) using partial knowledge of first principles of the underlying system (see at least Basso: ¶ [0047] & Figs. 1, 1A, 1B and 1C. Basso teaches that the crop model, in combination with complete historical weather information and partial historical yield information can be used to simulate the missing yield periods (e.g., with knowledge of what plant was planted during the missing yield periods).).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: wherein executing the optimizer comprises performing optimization using partial knowledge of first principles of the underlying system, and in further view of Basso, whereby a crop management plan is determined using a crop model incorporating a variety of inputs and plant-specific material and energy balances to specify one or more grower-controlled management parameters. An updated plan for a given field (e.g., reflecting an updated prescription for fertilization, irrigation, and/or other grower-controlled management parameters) can be followed by a grower to increase crop yield and/or optimize one or more other crop or field parameters (e.g., crop quality, field (marginal) net return, etc.) (see at least Basso: ¶ [0041].).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Basso, the results of the combination were predictable.
15. Claims 5, 19 and 33 are rejected under 35 U.S.C. 103 as being unpatentable over US PG Pub (US 2019/0057461 A1) hereinafter Ruff, et. al, in view of NPL Document: "Scheduling planting time through developing an optimization model and analysis of time series growing degree units." (2022), hereinafter Ansarifar, et. al., and in further view of US PG Pub (US 2009/0234695 A1) hereinafter Kapadi, et. al, and in further view of US PG Pub (US 2020/0090107 A1) hereinafter McKeeman, et. al.
Regarding Dependent Claims 5, 19 and 33, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does not explicitly disclose, but McKeeman in the analogous art for method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose the following:
- the underlying system (see at least McKeeman: ¶ [0107] & ¶ [0125] & Figs. 2A-2C. McKeeman notes the agricultural system 200 shown in Figs. 2A-2C.) comprises multiple resources who use is time-varying (see at least McKeeman: Figs. 2A-2C & ¶ [0065] & ¶ [0103]. McKeeman notes that producers may not own or have access to sufficient agricultural resources (for example, but not limited to, agricultural equipment, agricultural equipment operators, or agricultural equipment repair technicians), needed to service all of their needs, and the producers may be reliant on agricultural service providers. Agricultural service providers may provide agricultural equipment to the producer, and/or may provide agricultural equipment operators to the producer, and/or may provide repair services for agricultural equipment to the producer. See also McKeeman at ¶ [0005]: McKeeman notes that during critical windows of time during which services need to be completed, such as harvesting, where crop conditions and weather can change rapidly. See also McKeeman at ¶ [0103]: A scheduled service record may include a number of sub-fields comprising the information, such as (but not limited to) a producer identifier to whom the service is to be delivered, identifiers for one or more equipment owners scheduled to provide agricultural equipment for the service, identifiers for one or more operators scheduled to provide human resources to operate the service or to operate equipment for the service, one or more pieces of agricultural equipment that are to be used in providing the service, calendar times or days during which the service is to be performed or during which the equipment is to be hired, a date and time at which the booking was made, a booking identifier, pricing or payment information associated with the scheduled service, any conditions that must be met in order to trigger payment or a release of any escrowed funds.) and that are subject to one or more environmental factors (see at least McKeeman: ¶ [0065-0066] & ¶ [0174]. McKeeman notes that events that may impact a scheduled service may include (but are not limited to), a current weather condition, a change in a forecast weather condition, a transportation delay, an illness or other inability of a person to perform duties necessary for the service, an equipment breakdown, malfunction or need for maintenance, a theft of equipment. See also McKeeman at ¶ [0065-0066]: The demand for agricultural services varies greatly, depending on factors such as the seasons, weather patterns, the type of agricultural product for a particular producer (e.g., type of crop, type of livestock). There can be delays to the actual provision of services due to factors that are beyond the control of the producer, for example weather conditions, equipment transport delays, or lack of qualified agricultural equipment operators, since everything has to be simultaneously ready to deploy for the services to take place.);
- the optimizer (see at least McKeeman: Figs. 1C-1D & Figs. 2A-2C & ¶ [0063-0064]. McKeeman teaches that an agricultural management system is configured to optimize scheduling and facilitate asset tracking with a goal of maximizing the utilization of agricultural equipment and the agricultural equipment operators, while enabling producers to meet their needs in a timely manner with reliability and predictability.) determines a scheduling status for use of the multiple resources over time (see at least McKeeman: ¶ [0165-0167] & ¶ [0172] & ¶ [0262]. McKeeman teaches that an availability of equipment and/or operator resources compared against the service scheduling requirements or preferences of a producer. A location of equipment and/or operator resources compared against any location-based service requirements or preferences of a producer. See also McKeeman at ¶ [0165-0167]: Coordinate the booking and scheduling of a service amongst users of the system involved in providing or receiving the service, to update availabilities of equipment or human resources in accordance with a booked service, to reschedule services or to notify users in response to any change. Also having received a selected pairing 262 or any booking confirmation or payment, booking manager and scheduler 294 may mark, label or indicate the associated equipment and operator resources as only “tentatively available” (as opposed to available) within any of storages 271 in order to help the system detect any potential conflicts or problems of double-booking. Booking manager and scheduler 294 may temporarily mark, label or indicate the associated equipment or operator resources as reserved or unavailable within one or more of the appropriate storages 271 via the sending of suitable scheduling and availability update information 265 to the appropriate storage(s). See also McKeeman at ¶ [0172]: Booking manager and scheduler 294 may send further scheduling and availability update information 265 to any storage within storages 271 to mark, label or indicate the associated equipment or operator resources as available or unavailable in accordance with the current status of the booking, scheduling or payment processes, or to record the current status for example in scheduled services storage 219. See also McKeeman at Fig. 2A noting “status manager 214”.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: the underlying system comprises multiple resources who use is time-varying and that are subject to one or more environmental factors and the optimizer determines a scheduling status for use of the multiple resources over time, and in further view of McKeeman, in order to improve the utilization factor of agricultural machinery and equipment, and the human resources needed to operate such. The systems and methods enable seekers of an agricultural service (such as producers) to rapidly identify both equipment owners and operators and to contract out their use for predetermined scheduled periods as a farming service package. Moreover, the systems of McKeeman gather, track, store and analyze a plurality of different information types. The further analysis of this information may lead to many actionable uses, including real time scheduling of services, pairing service providers with one another to form service packages, pairing service providers and producers, finding the optimal service providers for a given producer in a search type format, providing real-time push notifications to available service providers when a producer need is identified, providing real-time push notifications to producers when there is service provider availability in their areas (see at least McKeeman: ¶ [0007-0008]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by McKeeman, the results of the combination were predictable.
16. Claims 10-11, 24-25 and 38-39 are rejected under 35 U.S.C. 103 as being unpatentable over US PG Pub (US 2019/0057461 A1) hereinafter Ruff, et. al, in view of NPL Document: "Scheduling planting time through developing an optimization model and analysis of time series growing degree units." (2022), hereinafter Ansarifar, et. al., and in further view of US PG Pub (US 2009/0234695 A1) hereinafter Kapadi, et. al, and in further view of US PG Pub (US 2019/0050948 A1) hereinafter Perry, et. al.
Regarding Dependent Claims 10, 24 and 38, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does not explicitly disclose, but Perry in the analogous art for method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose the following:
- wherein the predictions are generated using stochastic optimization (see at least Perry: ¶ [0142-0144]. Perry teaches that the Bayesian classifier can select a set of farming operations associated with a predicted crop production probability distribution that stochastically dominates the predicted crop production probability distributions associated with other possible sets of farming operations. See also at least Perry at ¶ [0144]: If a predicted probable yield range of the set of probable yield ranges stochastically dominates the predicted probable range for the planted crop, the multivariable regression model can then recommend a replant, and can identify a set of farming operations (including the type of crop to plant, planting date, and the like) that will optimize crop production for the replant.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: wherein the predictions are generated using stochastic optimization and in further view of Perry, whereby a crop prediction model can be a multivariable regression model trained on one or more of historic planting dates, planting rates, harvest dates, weather conditions, and solar radiation. If a predicted probable yield range of the set of probable yield ranges stochastically dominates the predicted probable range for the planted crop, the multivariable regression model can then recommend a replant, and can identify a set of farming operations (including the type of crop to plant, planting date, and the like) that will optimize crop production for the replant (see at least Perry: ¶ [0144]). Moreover, having access to predicted crop production information can beneficially enable growers to efficiently and profitably allocate resources, improve crop yield stability, reduce or account for short- and long-term risks, and evaluate expected market conditions months or years in advance (see at least Perry: ¶ [0186]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Perry, the results of the combination were predictable.
Regarding Dependent Claims 11, 25 and 39, Ruff / Ansarifar / Kapadi / Perry method / apparatus / non-transitory computer readable medium for machine learning based production optimizers teaches the limitations of Claims 1, 10, 15, 24 and 29 above, and Perry further teaches the method / apparatus / non-transitory computer readable medium for machine learning based production optimizers comprising:
- wherein the predictions are generated using reinforcement learning or Monte Carlo simulations (see at least Perry: ¶ [0007] & ¶ [0116] & ¶ [0176]. Perry teaches that various machine learning operations may be performed in different contexts to train the crop prediction models, and the crop prediction models can perform various machine learning operations when applied, including but not limited to: a generalized linear model, a generalized additive model, non-parametric regression, random forest, spatial regression, a Bayesian regression model, a time series analysis, a Bayesian network, a Gaussian network, decision tree learning, artificial neural networks, recurrent neural network, reinforcement learning, linear/non-linear regression, support vector machines, clustering operations, genetic algorithm operations, and any combination or order thereof. See also Perry at ¶ [0145]: The resulting Bayesian model can interpolate soil sample information over a portion of land, such as a grower's field, for instance using Markov Chain Monte Carlo sampling or variational inference estimations. See also Perry at ¶ [0176]: The crop prediction models can be learned with or without human guidance, and can be learned using supervised machine learning, unsupervised machine learning, and/or reinforcement machine learning. Examiner Note: Examiner interprets that the predictions in the Perry reference are primarily determined using “reinforcement learning” or “reinforcement machine learning”.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi / Perry method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: wherein the predictions are generated using reinforcement learning or Monte Carlo simulations and in further view of Perry, whereby a crop prediction model can be a multivariable regression model trained on one or more of historic planting dates, planting rates, harvest dates, weather conditions, and solar radiation. If a predicted probable yield range of the set of probable yield ranges stochastically dominates the predicted probable range for the planted crop, the multivariable regression model can then recommend a replant, and can identify a set of farming operations (including the type of crop to plant, planting date, and the like) that will optimize crop production for the replant (see at least Perry: ¶ [0144]). Moreover, having access to predicted crop production information can beneficially enable growers to efficiently and profitably allocate resources, improve crop yield stability, reduce or account for short- and long-term risks, and evaluate expected market conditions months or years in advance (see at least Perry: ¶ [0186]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Perry, the results of the combination were predictable.
17. Claims 14, 28 and 42 are rejected under 35 U.S.C. 103 as being unpatentable over US PG Pub (US 2019/0057461 A1) hereinafter Ruff, et. al, in view of NPL Document: "Scheduling planting time through developing an optimization model and analysis of time series growing degree units." (2022), hereinafter Ansarifar, et. al., and in further view of US PG Pub (US 2009/0234695 A1) hereinafter Kapadi, et. al, and in further view of US PG Pub (US 2017/0169523 A1) hereinafter Xu, et. al.
Regarding Dependent Claims 14, 28 and 42, Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does not explicitly disclose, but Xu in the analogous art for method / apparatus / non-transitory computer readable medium for machine learning based production optimizers does disclose the following:
- wherein generating the predictions comprises considering uncertainty in an objective function (see at least Xu: ¶ [0149] & ¶ [0163] & ¶ [0169]. Xu teaches that for example, agricultural intelligence computer system 130 may estimate total uncertainty in optimal nitrogen amounts, relative yield values, and total yield values based on model uncertainty, measurement uncertainty, and unknown weather situations. By estimating the uncertainty, agricultural intelligence computer system 130 may display probabilistic estimates for total yield as well as probabilistic recommendations for nitrogen applications. See also at least Xu at ¶ [0163]: The utility function typically represents a trade-off between exploration, using input values where the present knowledge of the objective function is very uncertain, and exploitation, using input values where the objective function is expected to be high. See also at least Xu at ¶ [0169]: The utility function is an information function that provides a numerical measure at every point x, of either the (1) uncertainty of the GP approximation of the objective function, or (2) likelihood that x is the maximum of the objective function, or (3) a combination of both.) used to generate the predictions (see at least Xu: ¶ [0095] & ¶ [0135] & ¶ [0149]. Xu teaches that the model data may include a model of past events on the one or more fields, a model of the current status of the one or more fields, and/or a model of predicted events on the one or more fields. See also Xu at ¶ [0095]: Cross validation may include comparison to ground truthing that compares predicted results with actual results on a field, such as a comparison of precipitation estimate with a rain gauge at the same location or an estimate of nitrogen content with a soil sample measurement. See also Xu at ¶ [0135]: By parameterizing the above equation for different portions of the crop's development, agricultural intelligence computer system 130 is better able to predict an effect on relative yield of nitrate in the soil at a particular time. Xu teaches that for example, agricultural intelligence computer system 130 may estimate total uncertainty in optimal nitrogen amounts, relative yield values, and total yield values based on model uncertainty, measurement uncertainty, and unknown weather situations. By estimating the uncertainty, agricultural intelligence computer system 130 may display probabilistic estimates for total yield as well as probabilistic recommendations for nitrogen applications.).
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to have combined / modified the teachings of Ruff / Ansarifar / Kapadi method / apparatus / non-transitory computer readable medium for machine learning based production optimizers with the aforementioned teachings of: wherein generating the predictions comprises considering uncertainty in an objective function, and in further view of Xu, in order for the agricultural intelligence computer system may estimate total uncertainty in optimal nitrogen amounts, relative yield values, and total yield values based on model uncertainty, measurement uncertainty, and unknown weather situations. By estimating the uncertainty, agricultural intelligence computer system may display probabilistic estimates for total yield as well as probabilistic recommendations for nitrogen applications. The probabilistic recommendations for nitrogen applications may allow farmers who are more risk averse to make more informed decisions (see at least Xu: ¶ [0149]).
Further, the claimed invention is merely a combination of old elements in a similar field for machine learning based production optimizers and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that, given the existing technical ability to combine the elements as evidenced by Xu, the results of the combination were predictable.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DERICK HOLZMACHER whose telephone number is (571) 270-7853. The examiner can normally be reached on Monday-Friday 9:00 AM – 6:30 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, Applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Brian Epstein can be reached on 571-270-5389. The fax phone number for the organization where this application or proceeding is assigned is 571-270-8853.
Information regarding the status of an application may be obtained from Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/DERICK J HOLZMACHER/Patent Examiner, Art Unit 3625A
/BRIAN M EPSTEIN/Supervisory Patent Examiner, Art Unit 3625
Prosecution Insights