DETAILED ACTION
Notice of AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
Applicant’s amendment and remarks dated 1/13/2026 have been considered. Claims 13-15 have been newly-added. Claims 1-8 and 13-15 are pending.
Response to Arguments
On page 5 of Applicant’s 1/13/2026 Amendment and remarks, Applicant asserts that paras. 0028-0032, 0035-0037, 0136-0140, and 0180 (of PGPUB US 20220229428 A1) provide written description support for the claim amendments and new claims.
The examiner agrees that the portions of the disclosure identified by Applicant provide sufficient written description support for the claim amendments and new claims 13-15.
On pages 5-6 of Applicant’s 1/13/2026 Amendment and remarks, with respect to the rejections under 35 U.S.C. 101, Applicant argues that as amended, the claims now integrate the judicial exception into a practical application in the “field of industrial manufacturing and production line optimization” and claims a “particular machine (a production line simulation system) configured in a specific way to solve the technological problem of determining necessary machine working times to meet shipping deadlines.”
The examiner respectfully disagrees. Each of the newly-added limitations are mental processes as explained in the detailed rejections below, and therefore are not available under Step 2A, Prong 2, to integrate the judicial exception into a practical application, or under Step 2B to demonstrate that the claims are “significantly more” than the judicial exception.
At most, the claims direct the mental processes to the particular field of use of calculating production line working times (including production times and shipping times), and MPEP 2106.04(d) explains that “Generally linking the use of a judicial exception to a particular technological environment or field of use” does not integrate a judicial exception into a practical application.
On pages 7-9 of Applicant’s 1/13/2026 Amendment and remarks, with respect to the rejections under 35 U.S.C. 103 to claims 1-5 and 8, Applicant argues that the newly-added limitations to independent claims 1 and 8 are not taught by the RUIZ, PAPANGELIS, and MUANDET references.
The examiner agrees that at least some of the limitations added to claims 1 and 8 are not taught or obvious in view of the RUIZ, PAPANGELIS, and/or MUANDET references. However, new grounds of rejection for the independent claims in view of the RUIZ and GARDEN references are made herein, where such new grounds of rejection are necessitated by Applicant’s amendments to claims 1 and 8.
On pages 9-10 of Applicant’s 1/13/2026 Amendment and remarks, with respect to the rejections under 35 U.S.C. 103 to claims 6-7, and with respect to new claims 13-15, Applicant argues that such claims should be allowed for the same reasons argued with respect to independent claims 1 and 8.
The examiner respectfully disagrees for the same reasons explained above with respect to independent claims 1 and 8.
Claim Rejections - 35 USC § 101
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.
Claims 1-8 and 13-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Step 1 of the Alice/Mayo framework, Claims 1-8 and 13-15 are each directed to a device (a machine), which falls within one of the four statutory categories of inventions.
Regarding Claim 1
Step 2A, prong 1 (Is the claim directed to a law of nature, a natural phenomenon or an abstract idea).
Claim 1 recites the following mental processes, that in each case under the broadest reasonable interpretation, covers performance of the limitation in the mind (including an observation, evaluation, judgment, opinion) or with the aid of pencil and paper but for the recitation of generic computer components (e.g., “memory”, “processor”, and “instructions”).
generate a third model indicating a first relationship between a first model and a parameter of a second model, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise a mental model (or write a model comprising of inputs, outputs, and logic to transform inputs to outputs down on a piece of paper), where such mental model shows the relationship between a first model and a parameter of a second model (e.g., the mental model shows a rule that the output of the first model and the input to a second model should be the same))
the first model indicating a second relationship between a sample and a label of the sample, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise a first model that indicates a relationship between a sample and a label of the sample, e.g., the relationship can be a mentally-estimated percentage of how well the label matches the sample)
the second model indicating the second relationship and being different from the first model (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the second model to be different than the first model, and to further relate to the association between the sample and the label)
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally correlate a sample corresponding to production amount of a product per unit time on a production line with a label corresponding to a shipping time for the production amount of the product per unit time)
wherein the parameter of the second model corresponds to a working time on the production line, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally have thew working time on the production line be a parameter of the second model)
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally configure the third model to have a label as input and an output corresponding to one of the parameters of the second model)
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally configure the second model to output a label corresponding to the label of the sample based on the sample and a parameter output by the third model)
Step 2A, prong 2 (Does the claim recite additional elements that integrate the judicial exception into a practical application?).
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements (e.g., “memory”, “processor”, and “instructions”) which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)).
Regarding the “model generation device comprising” limitation, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation attempts to cover a solution to an identified problem with no restriction on how the result is accomplished, or provides no description of the mechanism for accomplishing the result. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Regarding the “at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:” limitations, such limitations are recited at a high-level of generality and amount to no more than adding the words “apply it” (or an equivalent) with the judicial exception. In particular, the claim only recites the additional elements of memory, processors, and instructions. These additional elements are recited at a high-level of generality and amount to no more than mere instructions to apply the exception using generic computer components (memory, processors, and instructions). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Step 2B (Does the claim recite additional elements that amount to significantly more than the judicial exception?)
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional elements (e.g., “memory”, “processor”, and “instructions”) are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)).
Regarding the “model generation device comprising” limitation, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation attempts to cover a solution to an identified problem with no restriction on how the result is accomplished, or provides no description of the mechanism for accomplishing the result. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Regarding the “at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:” limitations, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation merely provides instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Regarding Claim 2
Step 2A, Prong 1
to generate the third model that receives input of a distribution of a function indicating the first model and outputs a distribution of a function indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the third model in a manner such that it receives inputs and provides outputs as recited in this limitation with respect to the first and second models)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 3
Step 2A, Prong 1
to generate the third model that receives input of a distribution of a function indicating the first model and outputs a point indicating a function indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the third model in a manner such that it receives inputs and provides outputs as recited in this limitation with respect to the first and second models)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 4
Step 2A, Prong 1
to generate the third model that receives input of a point indicating a function indicating the first model and outputs a distribution of a function indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the third model in a manner such that it receives inputs and provides outputs as recited in this limitation with respect to the first and second models)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 5
Step 2A, Prong 1
to generate the third model that receives input of a point indicating a function indicating the first model and outputs a point indicating a function indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the third model in a manner such that it receives inputs and provides outputs as recited in this limitation with respect to the first and second models)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 6
Step 2A, Prong 1
to generate, as the third model, a function of a reproducing Kernel Hilbert space (RKHS) that receives input of a kernel mean indicating the first model and outputs a kernel mean indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the third model in a manner such that the model implements a function withing a RKHS, such that the devices third model receives inputs and provides outputs in a kernel mean format as recited in this limitation with respect to the first and second models)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 7
Step 2A, Prong 1
to calculate a value of the parameter of the second model based on the kernel mean indicating the second model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally perform a calculation according to inputs/outputs and logic of a mentally-devised second model that takes as an input a kernel mean from the third model)
Regarding Step 2A, Prong 2, the claim does not include any additional elements that integrate the judicial exception into a practical application and regarding Step 2B, there are no additional elements recited that amount to significantly more than the judicial exception.
Regarding Claim 8
Step 2A, prong 1 (Is the claim directed to a law of nature, a natural phenomenon or an abstract idea).
Claim 8 recites the following mental processes, that in each case under the broadest reasonable interpretation, covers performance of the limitation in the mind (including an observation, evaluation, judgment, opinion) or with the aid of pencil and paper but for the recitation of generic computer components (e.g., “memory”, “processor”, and “instructions”).
calculate a parameter of a second model regarding a given sample of a first model by applying a third model to the given sample, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally calculate a parameter of a second mental model using a sample from a first mental model and by applying a third mental model)
the first model indicating a relationship between a sample and a label of the sample, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise a first model that indicates a relationship between a sample and a label of the sample, e.g., the relationship can be a mentally-estimated percentage of how well the label matches the sample)
the second model indicating the relationship and being different from the first model, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise the second model to be different than the first model, and to further relate to the association between the sample and the label)
the third model indicating a relationship between the first model and a parameter of the second model (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally devise a mental model (or write a model comprising of inputs, outputs, and logic to transform inputs to outputs down on a piece of paper), where such mental model shows the relationship between a first model and a parameter of a second model (e.g., the mental model shows a rule that the output of the first model and the input to a second model should be the same))
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally correlate a sample corresponding to production amount of a product per unit time on a production line with a label corresponding to a shipping time for the production amount of the product per unit time)
wherein the parameter of the second model corresponds to a working time on the production line, (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally have thew working time on the production line be a parameter of the second model)
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally configure the third model to have a label as input and an output corresponding to one of the parameters of the second model)
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally configure the second model to output a label corresponding to the label of the sample based on the sample and a parameter output by the third model)
Step 2A, prong 2 (Does the claim recite additional elements that integrate the judicial exception into a practical application?).
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements (e.g., “memory”, “processor”, and “instructions”) which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)).
Regarding the “parameter calculation device comprising” limitation, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation attempts to cover a solution to an identified problem with no restriction on how the result is accomplished, or provides no description of the mechanism for accomplishing the result. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Regarding the “at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:” limitations, such limitations are recited at a high-level of generality and amount to no more than adding the words “apply it” (or an equivalent) with the judicial exception. In particular, the claim only recites the additional elements of memory, processors, and instructions. These additional elements are recited at a high-level of generality and amount to no more than mere instructions to apply the exception using generic computer components (memory, processors, and instructions). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Step 2B (Does the claim recite additional elements that amount to significantly more than the judicial exception?)
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional elements (e.g., “memory”, “processor”, and “instructions”) are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)).
Regarding the “parameter calculation device comprising” limitation, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation attempts to cover a solution to an identified problem with no restriction on how the result is accomplished, or provides no description of the mechanism for accomplishing the result. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Regarding the “at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:” limitations, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation merely provides instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Regarding Claim 13
Step 2A, Prong 1
wherein the working time corresponds to a sum of a length of time required for an assembly process by the assembly apparatus and a length of time required for an inspection process by the inspection apparatus. (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally add the assembly time and the inspection time to derive a total working time)
Step 2A, Prong 2
Regarding the “wherein the production line includes an assembly apparatus and an inspection apparatus” limitation, such limitation amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use. As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not integrate a judicial exception into a practical application.
Step 2B
Regarding the “wherein the production line includes an assembly apparatus and an inspection apparatus” limitation, such limitation amounts to no more than generally linking the use of a judicial exception to a particular technological environment or field of use as explained above, which does not amount to significantly more than the judicial exception. MPEP 2106.05(h).
Regarding Claim 14
Step 2A, Prong 1
calculate the parameter of the second model using the generated third model (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally calculate a parameter of a second model using a third model)
determine a processing amount for each device in the production line based on the calculated parameter of the second model (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally determine a processing amount for each device in a production line based on calculated parameters of models)
Step 2A, Prong 2
Regarding the “generate a control signal for controlling the operation of the production line according to the determined processing amount” limitation, such limitation amounts to extra solution activity because it is a mere nominal or tangential addition to the claim, (see MPEP 2106.05(g)). Moreover, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception. In particular, the claim only recites the additional element of generating a control signal for undisclosed hardware. This additional element is recited at a high-level of generality and amounts to no more than mere instructions to apply the exception using a generic computer component. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Step 2B
Regarding the “generate a control signal for controlling the operation of the production line according to the determined processing amount” limitation, this limitation amounts to extra solution activity because it is a mere nominal or tangential addition to the claim, (see MPEP 2106.05(g)). Moreover, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation merely provides instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Regarding Claim 15
Step 2A, Prong 1
determine a processing amount for each device in the production line based on the calculated parameter of the second model (under the broadest reasonable interpretation, this limitation can be performed by a human mentally (or using a pencil and paper), for example, a human can mentally determine a processing amount for each device in a production line based on calculated parameters of models)
Step 2A, Prong 2
Regarding the “generate a control signal for controlling the operation of the production line according to the determined processing amount” limitation, such limitation amounts to extra solution activity because it is a mere nominal or tangential addition to the claim, (see MPEP 2106.05(g)). Moreover, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception. In particular, the claim only recites the additional element of generating a control signal for undisclosed hardware. This additional element is recited at a high-level of generality and amounts to no more than mere instructions to apply the exception using a generic computer component. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea (See MPEP 2106.05(f)).
Step 2B
Regarding the “generate a control signal for controlling the operation of the production line according to the determined processing amount” limitation, this limitation amounts to extra solution activity because it is a mere nominal or tangential addition to the claim, (see MPEP 2106.05(g)). Moreover, such limitation is recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, because the limitation merely provides instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Claim Rejections - 35 USC § 103
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 8, and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Ruiz, Nataniel, et al. "Learning to simulate." arXiv preprint arXiv:1810.02513 (May 14, 2019), hereinafter referenced as RUIZ, in view of US 20170290345 A1, hereinafter referenced as GARDEN.
Regarding Claim 1
RUIZ teaches:
A model generation device comprising: (RUIZ, p. 2, section 2.2: “Our goal is to generate a synthetic dataset such that the main task model (MTM) hθ, when trained on this dataset until convergence, achieves maximum accuracy on the test set.”
RUIZ, p. 5, section 4.2: “For the following experiments we use computer vision applications and thus require a generative scene model and an image rendering engine.”
Examiner’s Note: The device that trains the “main task model” corresponds to the recited “model generation device” where such device is a computer)
(RUIZ, p. 2, section 2.2: “Our goal is to generate a synthetic dataset such that the main task model (MTM) hθ, when trained on this dataset until convergence, achieves maximum accuracy on the test set. The test set is evidently not available during train time. Thus, the task of our algorithm is to maximize MTM’s performance on the validation set by generating suitable data. Similar to reinforcement learning, we define a policy πw parameterized by w that can sample parameters ψ ~ πw for the simulator. The simulator can be seen as a generative model G(x, y| ψ) which generates a set of data samples (x, y) conditioned on ψ.”
RUIZ, p. 3, Figure 1:
PNG
media_image1.png
154
522
media_image1.png
Greyscale
Examiner’s Note: the Simulator corresponds to the recited “first model”, the Main Task Model corresponds to the recited “second model”, and the Policy module corresponds to the recited “third model”, where the Policy module outputs a parameter (ψ) that relates to data that is used by the Simulator to generate a “set of data samples (x, y) conditioned on ψ” (corresponding to recited “first relationship between a first model and a parameter of a second model”) where the input set of data samples (x,y) corresponds to the recited “parameter of a second model”)
the first model indicating a second relationship between a sample and a label of the sample, (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”
RUIZ, p. 2, section 2.2: “The simulator can be seen as a generative model G(x, y| ψ) which generates a set of data samples (x, y) conditioned on ψ.”;
Examiner’s Note: the Simulator (corresponding to the recited “first model”) generates a sample (x) and its corresponding label (y) with the connection between x and y corresponding to the recited “second relationship”)
the second model indicating the second relationship and being different from the first model. (RUIZ, p. 3, section 2.2: “The reward R is computed as the negative loss L or some other accuracy metric on the validation set.”
Examiner’s Note: the reward (R) computed by the Main Task Model (corresponding to the recited “second model”) is a loss metric with respect to the input sample + label pair (x,y) (corresponding to recited “second relationship”) and the Main Task Model is a separate and different model than the Simulator)
However, RUIZ fails to explicitly teach:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:
generate a third model
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time,
wherein the parameter of the second model corresponds to a working time on the production line,
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model.
However, in view of MPEP 2144.04 VI.B, which states that duplication of parts “has no patentable significance unless a new and unexpected result is produced”, and MPEP 2144.04 VI.C, which states that the rearrangement of parts, particularly where such rearrangement does not modify the operation of the device, RUIZ makes obvious:
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”;
Examiner’s Note: the Policy module (corresponding to recited “third model”) as depicted in Fig. 1, is modified such that the data samples (x,y), where y is the label for data sample x, are passed-thru from the Main Task Model so that they are an input to the Policy module, and where the output of the Policy module (parameters ψ) are now adapted to correspond to such data samples (x,y) (which correspond to recited “parameter of the second model” as explained above; the examiner notes that passing-thru existing data through the MTM to the Policy module is simply the duplication of such data to provide as input to the Policy module (see MPEP 2144.04 VI.B) and a rearrangement of the data flows from the Simulator to the MTM to the Policy Module (see MPEP 2144.04 VI.C), which in each case is obvious as supported by the legal precedent of MPEP 2144.04 VI.)
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model. (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”;
Examiner’s Note: the Main Task Model (MTM) module (corresponding to recited “second model”) as depicted in Fig. 1, is modified such that the data samples (x,y), where y is the label for data sample x, are passed-thru and output from the MTM to the Policy module so that they are an input to the Policy module, where such label (y) is based on the sample (x) and is based on the output of the previous iteration of the Policy module output parameters; the examiner notes that passing-thru existing data through the MTM to the Policy module is simply the duplication of such data to provide as input to the Policy module (see MPEP 2144.04 VI.B) and a rearrangement of the data flows from the Simulator to the MTM to the Policy Module (see MPEP 2144.04 VI.C), which in each case is obvious as supported by the legal precedent of MPEP 2144.04 VI.)
One of ordinary skill would have been motivated to modify RUIZ (using its own teachings and the legal precedent of MPEP 2144.04 VI) in order to provide additional inputs to each of the Policy, Simulator, and MTM modules to increase the number of parameters considered by each module, and potentially to increase the accuracy of each module based on additional received information.
However, RUIZ does not teach or make obvious:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:
generate a third model
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time,
wherein the parameter of the second model corresponds to a working time on the production line,
However, in a related field of endeavor (machine learning techniques with respect to food assembly and delivering, see para. 0151), GARDEN teaches:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to: (GARDEN, para. 0256: “When logic is implemented as software and stored in memory, one skilled in the art will appreciate that logic or information, can be stored on any computer readable medium for use by or in connection with any computer and/or processor related system or method. In the context of this document, a memory is a computer readable medium that is an electronic, magnetic, optical, or other another physical device or means that contains or stores a computer and/or processor program. Logic and/or the information can be embodied in any computer readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions associated with logic and/or information.”;
Examiner’s Note: the RUIZ-GARDEN combination now utilizes the processor and memory of GARDEN to execute the policy module, simulator, and main task model of RUIZ)
generate a third model (GARDEN, para. 0151: “The images and/or ratings and/or rules can be used as training data for training the machine-learning system during a training period or training time. The system can use the trained examples during operation or runtime to produce patterns and paths based on blob analysis to achieve a desired distribution of sauce, cheese, and/or toppings for any given instance of pizza or other food item.”
Examiner’s Note: the policy model of RUIZ is now trained (corresponding to recited “generated”) using the training techniques of GARDEN)
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time, (GARDEN, para. 0008: “A processor-based system can dynamically generate, maintain, and update a dynamic order queue to sequence various orders for food items, and to control an assembly line and associated robots of the assembly line to assemble food items or food products per order.”;
GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108. Such estimated delivery times may take into account factors such as the complexity of preparation and the time required for the desired or defined cooking process associated with the ordered food item. Such estimated delivery times may also take into account factors such as road congestion, traffic, time of day, and other factors affecting the delivery of the food item by the order dispatch and en route cooking control systems 108. In other instances, the estimated delivery time may reflect the availability of the ordered food item on a delivery vehicle that has been pre-staged in a particular area.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, where the delivery time for the items (corresponding to recited “shipping time for the production amount of the product per unit time”) is based on the estimated time to produce the foot items in the production module plus the estimated cook time (corresponding to recited “production amount of a product per unit time”) with respect to the food assembly line (corresponding to recited “production line”) of GARDEN)
wherein the parameter of the second model corresponds to a working time on the production line, (GARDEN, para. 0008: “A processor-based system can dynamically generate, maintain, and update a dynamic order queue to sequence various orders for food items, and to control an assembly line and associated robots of the assembly line to assemble food items or food products per order.”;
GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108. Such estimated delivery times may take into account factors such as the complexity of preparation and the time required for the desired or defined cooking process associated with the ordered food item.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, where known cook times for items are input parameters to the Main Task Model (corresponding to recited “working time on the production line”))
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN as explained above. As disclosed by GARDEN, one of ordinary skill would have been motivated to do so in order to optimize a food assembly process so that “hot prepared food that is freshly cooked can be delivered to a consumer shortly after the conclusion of the cooking process.” (para. 0007). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models.
Regarding Claim 8
RUIZ teaches:
A parameter calculation device comprising: (RUIZ, p. 2, section 2.2: “Our goal is to generate a synthetic dataset such that the main task model (MTM) hθ, when trained on this dataset until convergence, achieves maximum accuracy on the test set.”
RUIZ, p. 5, section 4.2: “For the following experiments we use computer vision applications and thus require a generative scene model and an image rendering engine.”
Examiner’s Note: The device runs the MTM model corresponds to the recited “parameter calculation device” where such device is a computer)
calculate a parameter of a second model regarding a given sample of a first model by applying a third model to the given sample, (RUIZ, p. 2, section 2.2: “Our goal is to generate a synthetic dataset such that the main task model (MTM) hθ, when trained on this dataset until convergence, achieves maximum accuracy on the test set. The test set is evidently not available during train time. Thus, the task of our algorithm is to maximize MTM’s performance on the validation set by generating suitable data. Similar to reinforcement learning, we define a policy πw parameterized by w that can sample parameters ψ ~ πw for the simulator. The simulator can be seen as a generative model G(x, y| ψ) which generates a set of data samples (x, y) conditioned on ψ.”
RUIZ, p. 3, Figure 1:
PNG
media_image1.png
154
522
media_image1.png
Greyscale
Examiner’s Note: the Simulator corresponds to the recited “first model”, the Main Task Model corresponds to the recited “second model”, and the Policy module corresponds to the recited “third model”, where the Policy module calculates a parameter (ψ) that relates to data that is used by the Main Task Model which relates to the samples (x,y) of the Simulator)
the first model indicating a relationship between a sample and a label of the sample, (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”
RUIZ, p. 2, section 2.2: “The simulator can be seen as a generative model G(x, y| ψ) which generates a set of data samples (x, y) conditioned on ψ.”;
Examiner’s Note: the Simulator (corresponding to the recited “first model”) generates a sample (x) and its corresponding label (y) with the connection between x and y corresponding to the recited “relationship”)
the second model indicating the relationship and being different from the first model, (RUIZ, p. 3, section 2.2: “The reward R is computed as the negative loss L or some other accuracy metric on the validation set.”
Examiner’s Note: the reward (R) computed by the Main Task Model (corresponding to the recited “second model”) is a loss metric with respect to the input sample + label pair (x,y) (corresponding to recited “relationship”) and the Main Task Model is a separate and different model than the Simulator)
the third model indicating a relationship between the first model and a parameter of the second model. (RUIZ, p. 2, section 2.2: “Our goal is to generate a synthetic dataset such that the main task model (MTM) hθ, when trained on this dataset until convergence, achieves maximum accuracy on the test set. The test set is evidently not available during train time. Thus, the task of our algorithm is to maximize MTM’s performance on the validation set by generating suitable data. Similar to reinforcement learning, we define a policy πw parameterized by w that can sample parameters ψ ~ πw for the simulator. The simulator can be seen as a generative model G(x, y| ψ) which generates a set of data samples (x, y) conditioned on ψ.”
RUIZ, p. 3, Figure 1:
PNG
media_image1.png
154
522
media_image1.png
Greyscale
Examiner’s Note: the Simulator corresponds to the recited “first model”, the Main Task Model corresponds to the recited “second model”, and the Policy module corresponds to the recited “third model”, where the Policy module outputs a parameter (ψ) that relates to data that is used by the Simulator to generate a “set of data samples (x, y) conditioned on ψ” (corresponding to recited “relationship between a first model and a parameter of a second model”) where the input set of data samples (x,y) corresponds to the recited “parameter of a second model”)
However, RUIZ fails to explicitly teach:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time,
wherein the parameter of the second model corresponds to a working time on the production line,
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model.
However, in view of MPEP 2144.04 VI.B, which states that duplication of parts “has no patentable significance unless a new and unexpected result is produced”, and MPEP 2144.04 VI.C, which states that the rearrangement of parts, particularly where such rearrangement does not modify the operation of the device, RUIZ makes obvious:
wherein the third model is configured to acquire the label as an input, and output a parameter corresponding to the parameter of the second model, and (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”;
Examiner’s Note: the Policy module (corresponding to recited “third model”) as depicted in Fig. 1, is modified such that the data samples (x,y), where y is the label for data sample x, are passed-thru from the Main Task Model so that they are an input to the Policy module, and where the output of the Policy module (parameters ψ) are now adapted to correspond to such data samples (x,y) (which correspond to recited “parameter of the second model” as explained above; the examiner notes that passing-thru existing data through the MTM to the Policy module is simply the duplication of such data to provide as input to the Policy module (see MPEP 2144.04 VI.B) and a rearrangement of the data flows from the Simulator to the MTM to the Policy Module (see MPEP 2144.04 VI.C), which in each case is obvious as supported by the legal precedent of MPEP 2144.04 VI.)
wherein the second model is configured to output a label corresponding to the label of the sample based on the sample and the parameter output by the third model. (RUIZ, p. 1, section 1: “Data x usually arises from a real world process (for instance, someone takes a picture with a camera) and labels y are often annotated by humans (someone describing the content of that picture).”;
Examiner’s Note: the Main Task Model (MTM) module (corresponding to recited “second model”) as depicted in Fig. 1, is modified such that the data samples (x,y), where y is the label for data sample x, are passed-thru and output from the MTM to the Policy module so that they are an input to the Policy module, where such label (y) is based on the sample (x) and is based on the output of the previous iteration of the Policy module output parameters; the examiner notes that passing-thru existing data through the MTM to the Policy module is simply the duplication of such data to provide as input to the Policy module (see MPEP 2144.04 VI.B) and a rearrangement of the data flows from the Simulator to the MTM to the Policy Module (see MPEP 2144.04 VI.C), which in each case is obvious as supported by the legal precedent of MPEP 2144.04 VI.)
One of ordinary skill would have been motivated to modify RUIZ (using its own teachings and the legal precedent of MPEP 2144.04 VI) in order to provide additional inputs to each of the Policy, Simulator, and MTM modules to increase the number of parameters considered by each module, and potentially to increase the accuracy of each module based on additional received information.
However, RUIZ does not teach or make obvious:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to:
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time,
wherein the parameter of the second model corresponds to a working time on the production line,
However, in a related field of endeavor (machine learning techniques with respect to food assembly and delivering, see para. 0151), GARDEN teaches:
at least one memory configured to store the instructions; and at least one processor configured to execute the instructions to: (GARDEN, para. 0256: “When logic is implemented as software and stored in memory, one skilled in the art will appreciate that logic or information, can be stored on any computer readable medium for use by or in connection with any computer and/or processor related system or method. In the context of this document, a memory is a computer readable medium that is an electronic, magnetic, optical, or other another physical device or means that contains or stores a computer and/or processor program. Logic and/or the information can be embodied in any computer readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions associated with logic and/or information.”;
Examiner’s Note: the RUIZ-GARDEN combination now utilizes the processor and memory of GARDEN to execute the policy module, simulator, and main task model of RUIZ)
wherein the sample corresponds to a production amount of a product per unit time on a production line, and the label corresponds to a shipping time for the production amount of the product per unit time, (GARDEN, para. 0008: “A processor-based system can dynamically generate, maintain, and update a dynamic order queue to sequence various orders for food items, and to control an assembly line and associated robots of the assembly line to assemble food items or food products per order.”;
GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108. Such estimated delivery times may take into account factors such as the complexity of preparation and the time required for the desired or defined cooking process associated with the ordered food item. Such estimated delivery times may also take into account factors such as road congestion, traffic, time of day, and other factors affecting the delivery of the food item by the order dispatch and en route cooking control systems 108. In other instances, the estimated delivery time may reflect the availability of the ordered food item on a delivery vehicle that has been pre-staged in a particular area.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, where the delivery time for the items (corresponding to recited “shipping time for the production amount of the product per unit time”) is based on the estimated time to produce the foot items in the production module plus the estimated cook time (corresponding to recited “production amount of a product per unit time”) with respect to the food assembly line (corresponding to recited “production line”) of GARDEN)
wherein the parameter of the second model corresponds to a working time on the production line, (GARDEN, para. 0008: “A processor-based system can dynamically generate, maintain, and update a dynamic order queue to sequence various orders for food items, and to control an assembly line and associated robots of the assembly line to assemble food items or food products per order.”;
GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108. Such estimated delivery times may take into account factors such as the complexity of preparation and the time required for the desired or defined cooking process associated with the ordered food item.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, where known cook times for items are input parameters to the Main Task Model (corresponding to recited “working time on the production line”))
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN as explained above. As disclosed by GARDEN, one of ordinary skill would have been motivated to do so in order to optimize a food assembly process so that “hot prepared food that is freshly cooked can be delivered to a consumer shortly after the conclusion of the cooking process.” (para. 0007). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models.
Regarding Claim 14
RUIZ and GARDEN teach the device according to claim 1. RUIZ further makes obvious:
calculate the parameter of the second model using the generated third model; (Examiner’s Note: the Policy module (corresponding to recited “generated third model”) as depicted in Fig. 1, is modified such that output is passed directly as input to the MTM module (the recited “second model”); the examiner notes that passing existing data through the Policy Module to the MTM module is simply the duplication of such data to provide as input to the MTM module (see MPEP 2144.04 VI.B) and a rearrangement of the data flows from the Policy Module to the Simulator to the MTM (see MPEP 2144.04 VI.C), which in each case is obvious as supported by the legal precedent of MPEP 2144.04 VI.)
One of ordinary skill would have been motivated to modify RUIZ (using its own teachings and the legal precedent of MPEP 2144.04 VI) in order to provide additional inputs to each of the Policy, Simulator, and MTM modules to increase the number of parameters considered by each module, and potentially to increase the accuracy of each module based on additional received information.
However, RUIZ does not teach or make obvious:
determine a processing amount for each device in the production line based on the calculated parameter of the second model; and
generate a control signal for controlling the operation of the production line according to the determined processing amount.
However, in a related field of endeavor (machine learning techniques with respect to food assembly and delivering, see para. 0151), GARDEN teaches:
determine a processing amount for each device in the production line based on the calculated parameter of the second model; and (GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, to estimate the processing amount times for the different steps of the assembly line (including cooking time) based on the overall delivery time calculated by the revised Main Task Model)
generate a control signal for controlling the operation of the production line according to the determined processing amount. (GARDEN, para. 0013: “generating, by the control system, control signals based on the respective orders for food items, and conveying, by a conveyor, a plurality of instances of the food items along at least a portion of the robotic food preparation assembly line; and causing, by the control system, a respective tool of a respective appendage of each of a plurality of robots to assemble the instances of the food items based at least in part on the control signals,”
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to generate control signals for operating the different devices of the assembly line of GARDEN)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN as explained above. As disclosed by GARDEN, one of ordinary skill would have been motivated to do so in order to optimize a food assembly process so that “hot prepared food that is freshly cooked can be delivered to a consumer shortly after the conclusion of the cooking process.” (para. 0007).
Regarding Claim 15
RUIZ and GARDEN teach the device according to claim 8. However, RUIZ does not teach or make obvious:
determine a processing amount for each device in the production line based on the calculated parameter of the second model; and
generate a control signal for controlling the operation of the production line according to the determined processing amount.
However, in a related field of endeavor (machine learning techniques with respect to food assembly and delivering, see para. 0151), GARDEN teaches:
determine a processing amount for each device in the production line based on the calculated parameter of the second model; and (GARDEN, para. 0182: “In at least some instances, the order front end server computer control systems 104 can provide the consumer placing an order for a food item with an estimated delivery time for the item. In at least some instances, the estimated delivery time may be based on the time to produce the food item in the production module plus the estimated time to cook the food item in transit by the order dispatch and en route cooking control systems 108.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to perform the estimations of GARDEN, to estimate the processing amount times for the different steps of the assembly line (including cooking time) based on the overall delivery time calculated by the revised Main Task Model)
generate a control signal for controlling the operation of the production line according to the determined processing amount. (GARDEN, para. 0013: “generating, by the control system, control signals based on the respective orders for food items, and conveying, by a conveyor, a plurality of instances of the food items along at least a portion of the robotic food preparation assembly line; and causing, by the control system, a respective tool of a respective appendage of each of a plurality of robots to assemble the instances of the food items based at least in part on the control signals,”
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to generate control signals for operating the different devices of the assembly line of GARDEN)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN as explained above. As disclosed by GARDEN, one of ordinary skill would have been motivated to do so in order to optimize a food assembly process so that “hot prepared food that is freshly cooked can be delivered to a consumer shortly after the conclusion of the cooking process.” (para. 0007).
Claims 2-5 are rejected under 35 U.S.C. 103 as being unpatentable over RUIZ in view of GARDEN and further in view of US 20180233143 A1, hereinafter referenced as PAPANGELIS.
Regarding Claim 2
RUIZ and GARDEN teach the device according to claim 1. RUIZ further teaches:
the third model that receives input of a distribution of a function indicating the first model (RUIZ, p. 5, section 4.1: “The top row of figure 2 illustrates how the policy gradually adjusts the data generating distribution q(x, y | ψ ) such that reward R is increased.”;
Examiner’s Note: RUIZ teaches that the Policy module (corresponding to the recited “third model”) receives as input the data generating distribution related to the Simulator (corresponding to recited “indicating the first model” limitation))
However, RUIZ and GARDEN fail to explicitly teach:
outputs a distribution of a function indicating the second model.
However, in a related field of endeavor (trained policy models, see para. 0004), PAPANGELIS teaches:
outputs a distribution of a function indicating the second model. (PAPANGELIS, para. 0096: “the policy model generates probability distributions for the domain-specific inputs (slots), using the domain independent parameters.”;
Examiner’s Note: the RUIZ-GARDEN-PAPANGELIS combination now modifies the Policy module of RUIZ so that it also generates a probability distribution with respect to the Main Task Model, such as a probability distribution relating to the accuracy of the Main Task Model)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and PAPANGELIS as explained above. As disclosed by PAPANGELIS, one of ordinary skill would have been motivate to do so in order to train a model to determine a next system action according to a policy. (para. 0078). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models, and the benefit of having a probability distribution as an output instead of a single value or variable (e.g., to show probabilities associated with different potential outcomes).
Regarding Claim 3
RUIZ and GARDEN teach the device according to claim 1. RUIZ further teaches:
the third model that receives input of a distribution of a function indicating the first model (RUIZ, p. 5, section 4.1: “The top row of figure 2 illustrates how the policy gradually adjusts the data generating distribution q(x, y | ψ ) such that reward R is increased.”;
Examiner’s Note: RUIZ teaches that the Policy module (corresponding to the recited “third model”) receives as input the data generating distribution related to the Simulator (corresponding to recited “indicating the first model” limitation))
However, RUIZ fails to explicitly teach:
outputs a point indicating a function indicating the second model.
However, in a related field of endeavor (trained policy models, see para. 0004), PAPANGELIS teaches:
outputs a point indicating a function indicating the second model. (PAPANGELIS, para. 0093: “Thus a list of actions, comprising action functions combined with parameterised action function inputs is inputted into the policy model, and the policy model selects and outputs one of these actions, comprising the action function combined with the parameterised input.”;
Examiner’s Note: the RUIZ-GARDEN-PAPANGELIS combination now modifies the Policy module of RUIZ so that it outputs a single value according to a function that creates a plurality of values with respect to the Main Task Model, such as a the reward value, calculated using a reward function with respect to the accuracy of the Main Task Model)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and PAPANGELIS as explained above. As disclosed by PAPANGELIS, one of ordinary skill would have been motivate to do so in order to train a model to determine a next system action according to a policy. (para. 0078). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models, and the benefit of having a single value as the output of the policy model, in addition to (or in lieu of) a probability distribution output.
Regarding Claim 4
RUIZ and GARDEN teach the device according to claim 1. RUIZ further teaches:
the third model that receives input of a point indicating a function indicating the first model (RUIZ, p. 3, Figure 1:
PNG
media_image1.png
154
522
media_image1.png
Greyscale
Examiner’s Note: RUIZ discloses receiving a point value (R) that is indicative of a reward function based on a loss metric with respect to the difference between a data point (x,y) sampled from the Simulator)
However, RUIZ fails to explicitly teach:
outputs a distribution of a function indicating the second model.
However, in a related field of endeavor (trained policy models, see para. 0004), PAPANGELIS teaches:
outputs a distribution of a function indicating the second model. (PAPANGELIS, para. 0096: “the policy model generates probability distributions for the domain-specific inputs (slots), using the domain independent parameters.”;
Examiner’s Note: the RUIZ-GARDEN-PAPANGELIS combination now modifies the Policy module of RUIZ so that it also generates a probability distribution with respect to the Main Task Model, such as a probability distribution relating to the accuracy of the Main Task Model)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and PAPANGELIS as explained above. As disclosed by PAPANGELIS, one of ordinary skill would have been motivate to do so in order to train a model to determine a next system action according to a policy. (para. 0078). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models, and the benefit of having a probability distribution as an output instead of a single value or variable (e.g., to show probabilities associated with different potential outcomes).
Regarding Claim 5
RUIZ and GARDEN teach the device according to claim 1. RUIZ further teaches:
the third model that receives input of a point indicating a function indicating the first model (RUIZ, p. 3, Figure 1:
PNG
media_image1.png
154
522
media_image1.png
Greyscale
Examiner’s Note: RUIZ discloses receiving a point value (R) that is indicative of a reward function based on a loss metric with respect to the difference between a data point (x,y) sampled from the Simulator)
However, RUIZ fails to explicitly teach:
outputs a point indicating a function indicating the second model.
However, in a related field of endeavor (trained policy models, see para. 0004), PAPANGELIS teaches:
and outputs a point indicating a function indicating the second model. (PAPANGELIS, para. 0093: “Thus a list of actions, comprising action functions combined with parameterised action function inputs is inputted into the policy model, and the policy model selects and outputs one of these actions, comprising the action function combined with the parameterised input.”;
Examiner’s Note: the RUIZ-GARDEN-PAPANGELIS combination now modifies the Policy module of RUIZ so that it outputs a single value according to a function that creates a plurality of values with respect to the Main Task Model, such as a the reward value, calculated using a reward function with respect to the accuracy of the Main Task Model)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and PAPANGELIS as explained above. As disclosed by PAPANGELIS, one of ordinary skill would have been motivate to do so in order to train a model to determine a next system action according to a policy. (para. 0078). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models, and the benefit of having a single value as the output of the policy model, in addition to (or in lieu of) a probability distribution output.
Claims 6-7 are rejected under 35 U.S.C. 103 as being unpatentable over RUIZ in view of GARDEN and further in view of Muandet, Krikamol, et al. "Kernel mean estimation and Stein effect." International Conference on Machine Learning. PMLR, 2014, hereinafter referenced as MUANDET.
Regarding Claim 6
RUIZ and GARDEN teach the device according to claim 1. However, RUIZ and GARDEN fail to explicitly teach:
wherein the at least one processor is configured to execute the instructions to generate, as the third model, a function of a reproducing Kernel Hilbert space (RKHS) that receives input of a kernel mean indicating the first model and outputs a kernel mean indicating the second model.
However, in a related field of endeavor (machine learning, see p. 1, section 1), MUANDET teaches:
wherein the at least one processor is configured to execute the instructions to generate, as the third model, a function of a reproducing Kernel Hilbert space (RKHS) that receives input of a kernel mean indicating the first model and outputs a kernel mean indicating the second model. (MUANDET, p. 1, section 1: “This paper aims to improve the estimation of the mean function in a reproducing kernel Hilbert space (RKHS) from a finite sample. A kernel mean of a probability distribution P over a measurable space X is defined by ....”;
Examiner’s Note: the RUIZ-GARDEN-MUANDET combination now modifies the models of RUIZ, as automated using the computer-implemented teachings of GARDEN, to utilize the teachings of MUANDET so that the input and output of the Policy model of Ruiz (corresponding to recited “third model”) is now a kernel mean function of a RKHS (derived from a sample point) as taught by MUANDET, where the input kernel mean relates to the Simulator and the output kernel mean relates to the Main Task Model)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and MUANDET as explained above. As disclosed by MUANDET, one of ordinary skill would have been motivate to do so because the “kernel mean has recently gained attention in the machine learning community, thanks to the introduction of Hilbert space embedding for distributions. ... Representing the distribution as a mean function in the RKHS has several advantages: 1) the presentation with appropriate choice of kernel k has been show to preserve all information about the distribution ...; 2) basic operations on the distribution can be carried out by means of inner products in RKHS ...; 3) no intermediate density estimation is required....” (p. 1, section 1).
Regarding Claim 7
RUIZ and GARDEN and MUANDET teach the device according to claim 6. However, RUIZ and GARDEN fail to explicitly teach:
wherein the at least one processor is configured to execute the instructions to calculate a value of the parameter of the second model based on the kernel mean indicating the second model.
However, in a related field of endeavor (machine learning, see p. 1, section 1), MUANDET teaches:
wherein the at least one processor is configured to execute the instructions to calculate a value of the parameter of the second model based on the kernel mean indicating the second model. (MUANDET, p. 1, section 1: “This paper aims to improve the estimation of the mean function in a reproducing kernel Hilbert space (RKHS) from a finite sample. A kernel mean of a probability distribution P over a measurable space X is defined by ....”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the models of RUIZ, as automated using the computer-implemented teachings of GARDEN, to utilize the teachings of MUANDET so that the output of the Policy model of Ruiz (corresponding to recited “third model”) is now a kernel mean function of a RKHS (derived from a sample point) as taught by MUANDET, where the output kernel mean relates to the Main Task Model which now calculates an output parameter based on input of a kernel mean)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and MUANDET as explained above. As disclosed by MUANDET, one of ordinary skill would have been motivate to do so because the “kernel mean has recently gained attention in the machine learning community, thanks to the introduction of Hilbert space embedding for distributions. ... Representing the distribution as a mean function in the RKHS has several advantages: 1) the presentation with appropriate choice of kernel k has been show to preserve all information about the distribution ...; 2) basic operations on the distribution can be carried out by means of inner products in RKHS ...; 3) no intermediate density estimation is required....” (p. 1, section 1).
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over RUIZ in view of GARDEN and further in view of US 4519300 A, hereinafter referenced as ADOMIS.
Regarding Claim 13
RUIZ and GARDEN teach the device according to claim 1. However, RUIZ fails to explicitly teach:
wherein the production line includes an assembly apparatus and an inspection apparatus, and
wherein the working time corresponds to a sum of a length of time required for an assembly process by the assembly apparatus and a length of time required for an inspection process by the inspection apparatus.
However, in a related field of endeavor (machine learning techniques with respect to food assembly and delivering, see para. 0151), GARDEN teaches:
wherein the production line includes an assembly apparatus and an inspection apparatus, and (GARDEN, para. 0080: “The on-demand robotic food assembly line 102 may include a first or primary assembly conveyor 122a.”;
GARDEN, para. 0108: “The refrigeration station 161 may provide for monitoring of the one or more workstations 124 enclosed within the refrigerated environment. For example, one or more windows 165 may provide for visual inspection, either by an operation and/or by an automated visual inspection system, of the interior of the refrigeration station 161.”;
Examiner’s Note: the RUIZ-GARDEN combination now modifies the Main Task Model of RUIZ to the kitchen assembly environment of GARDEN, where GARDEN teaches that the assembly line includes at least a conveyer (corresponding to “assembly apparatus”) and a visual inspection system (corresponding to recited “inspection apparatus”))
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN as explained above. As disclosed by GARDEN, one of ordinary skill would have been motivated to do so in order to optimize a food assembly process so that “hot prepared food that is freshly cooked can be delivered to a consumer shortly after the conclusion of the cooking process.” (para. 0007). One of ordinary skill would further understand the benefit of using a generic processor, memory, and computer instructions to automate models.
However, RUIZ and GARDEN fail to explicitly teach:
wherein the working time corresponds to a sum of a length of time required for an assembly process by the assembly apparatus and a length of time required for an inspection process by the inspection apparatus.
However, in a related field of endeavor (manufacturing processes), ADOMIS teaches:
wherein the working time corresponds to a sum of a length of time required for an assembly process by the assembly apparatus and a length of time required for an inspection process by the inspection apparatus. (ADOMIS, col. 1, lines 37-43: “Because of the above problem, it is necessary to visually inspect the passage in every piston with a cystoscope type instrument to insure that there are no cracks or cavities in the passage. Such 100% inspection adds to the manufacturing time and thus increases the total cost of manufacturing the piston.;
Examiner’s Note: ADOMIS teaches that the inspection time is added to the manufacturing time; the RUIZ-GARDEN-ADOMIS combination now calculates the total time for cooking and inspecting the results when determining the delivery times of GARDEN)
Before the effective filing date of the present application, it would have been obvious to one of ordinary skill in the art to modify the approach of RUIZ, that includes a Policy module, a Simulator, and a Main Task Model, with the teachings of GARDEN and ADOMIS as explained above. As disclosed by ADOMIS, one of ordinary skill would have been motivated to do so in order to add a visual inspection time to the finished product to ensure that there are no problems. (ADOMIS, col. 1, lines 37-43).
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.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 20190295034 A1 (Wenzel). “In some embodiments, the lead time includes a production time that indicates an amount of time required for the equipment supplier to produce or obtain the quantity of each of the one or more equipment units. The lead time includes a shipping time that indicates a second amount of time required to transport the one or more equipment units from the equipment supplier to a location of the building equipment, according to some embodiments.” (para. 0022).
Zhang, Hao, et al. "A digital twin-based approach for designing and multi-objective optimization of hollow glass production line." Ieee Access 5 (2017): 26901-26911. “the digital twin-based platform can simulate production order and accurate predict the order delivery time, through collecting production takt and calculating production load.” (p. 26909, section VI.A).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHAEL C LEE whose telephone number is (571)272-4933. The examiner can normally be reached M-F 12:00 pm - 8:00 pm ET.
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, Omar Fernandez Rivas can be reached at 571-272-2589. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MICHAEL C. LEE/Examiner, Art Unit 2128
/LUIS A SITIRICHE/Primary Examiner, Art Unit 2126