DETAILED ACTION
NOTICE OF PRE-AIA OR 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 .
STATUS OF CLAIMS
This action is in response to the Applicant’s filing on 11/08/2024. Claims 1-11 are pending and are examined below.
PRIORITY
Acknowledgement is made of Applicant’s claim of foreign priority to DE102023211071.7, filed on 11/08/2023.
CLAIM OBJECTIONS
Claims 5, 6 and 8 are objected to because of claim informalities.
As to claim 5, the second instance of “operating parameter range” should be amended to “the operating parameter range” to follow established antecedent basis.
As to claim 6, the element “deactivating same” should be amended to “deactivating [[same]] the further machine learning model” for formality and clarity.
As to claim 8:
In the element “to also combine,” Examiner suggests deleting the word “also” as it is redundant and adds confusion to the claim. If Applicant’s intention is to specify that the preceding steps are further performed by the controller instead of the machine learning models, the claim should read: “wherein the controller is further configured to [[also]] combine.” If instead Applicant’s intention is that the machine learning model is to perform the “combine” and “determine” steps, then Examiner respectfully requests further clarification on this front in the form of remarks and/or amendments.
“said sensor” lacks antecedent basis – Examiner suggests amending to “[[said]] the steering variable sensor”
Appropriate correction is required.
CLAIM REJECTIONS—35 U.S.C. § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 8-11 are rejected under 35 U.S.C. § 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. § 112, the applicant), regards as the invention.
As to claim 8, the recitation “a controller configured to … provide at least two trained machine learning models, wherein the machine learning models are trained to detect a hands-off state based on at least the captured steering variable and to output the hands-off state as output data, to supply the captured steering variable as input data to the trained machine learning models, to also combine the output data in weighted form to form a hands-off state and to provide them, and to determine values of the weightings based on a current context” (emphases added) is vague and indefinite. Namely, based on how the claim is written it is unclear whether the “combine” and “determine” steps are performed by the controller or by the machine learning models. The specification does not appear to explain which component performs the respective steps. Accordingly, it is unclear what Applicant is attempting to claim.
In light of the above, it is unclear what is being claimed in light of Applicant’s original disclosure.
Claims 9-11 depend from claim 8.
Therefore, claims 8-11 are rejected under 35 U.S.C. § 112(b) or 35 U.S.C. § 112 (pre-AIA ), second paragraph.
Appropriate correction is required.
CLAIM REJECTIONS—35 U.S.C. § 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.
Claim(s) 1-11 is/are rejected under 35 U.S.C. § 101 because the claims fail to pass the Alice/Mayo test for determining patent eligibility.
The patent eligibility test is performed below for independent claims 1 and 8.
Step 1—Does the claim fall within a statutory category?
Claim 1: Yes, the claim recites a process.
Claim 8: Yes, the claim recites a machine or manufacture.
Step 2A, Prong One—Is a judicial exception recited?
Claims 1 and 8 are provided below with the abstract idea indicated in bold and additional elements without bold.
1. A method for detecting a hands-off state at a steering wheel, the method which comprises:
capturing a steering variable at the steering wheel;
supplying the steering variable as input data to at least two trained machine learning models, wherein the machine learning models are trained to detect a hands-off state on a basis of at least the captured steering variable and to output corresponding output data;
determining weighting values on a basis of a current context; and
combining the output data in weighted form to form and provide a hands-off state.
8. An apparatus for detecting a hands-off state at a steering wheel, the apparatus comprising:
a steering variable sensor configured to capture a steering variable at the steering wheel;
a controller configured to receive the steering variable captured by said sensor, to provide at least two trained machine learning models, wherein the machine learning models are trained to detect a hands-off state based on at least the captured steering variable and to output the hands-off state as output data, to supply the captured steering variable as input data to the trained machine learning models, to also combine the output data in weighted form to form a hands-off state and to provide them, and to determine values of the weightings based on a current context.
The above shows: yes, a judicial exception is recited. But for the additional elements, the claim limitation pertaining to detecting a hands-off state, determining weighting values, and combining output data in weighted form to provide a hands-off state are processes which can practically be performed in the human mind with or without the use of a physical aid. Specifically, the broadest reasonable interpretation (BRI) of the claim encompasses performing evaluations and judgments over obtained data. The courts have held such forms of observation, evaluation, judgment, or opinion to represent the abstract idea of a mental process. As a result, the bolded limitations represent a mental process. Hence, the claim recites an abstract idea. (See MPEP § 2106.04(a)(2)(C)(III).)
Step 2A, Prong Two—Is the abstract idea integrated into a practical application?
No. The claims as a whole merely use generic computer components — i.e., a controller — that are recited at a high level of generality such that they cannot be considered more than mere instructions to apply the judicial exception using generic computer components. Therefore, the abstract idea is not integrated into a practical application.
Furthermore, the claimed “machine learning models” and associated limitations do not integrate the abstract idea into a practical application because the claim is merely using the machine learning models in a generic fashion to apply the abstract idea in a certain technological environment (i.e., machine learning) without putting forth an improvement in how the machine learning models function. Outside of nominal mentions of ordinary features and functions of generic machine learning models, the claimed invention fails to set forth sufficient detail as to how the machine learning models accomplish their claimed features in a way that differs from performing the operations through ordinary features of generic machine learning models. Contrast with Ex Parte Desjardins1, wherein the claimed invention was found eligible as it provides identified improvements as to how the machine learning model itself operates (e.g., training a machine learning model to learn new tasks while protecting knowledge about previous tasks to overcome the problem of “catastrophic forgetting” encountered in continual learning systems). Hence, Applicant’s claimed machine learning models merely indicate a field of use or technological environment in which the judicial exception is performed. That is, the claim merely confines the use of the abstract idea to a particular technological environment (machine learning) and therefore fails to integrate the claims into a practical application. (See MPEP 2106.05(h).)
Step 2B—Does the claim provide an inventive concept?
No. The additional elements of the claims amount to either:
Insignificant pre-solution activity in the form of mere data gathering:
capturing a steering variable at a steering wheel
supplying the steering variable as input data
a steering variable sensor configured to capture a steering variable at a steering wheel
Here, the “steering variable sensor” is a generic computing component used to set out insignificant activity; such fails to provide an inventive concept.
receive the steering variable captured by the sensor
Insignificant post-solution activity in the form of well-understood and conventional activity:
output corresponding output data
output the hands-off state as output data
Claims 2-7 depend from claim 1 but do not render the claimed invention patent eligible because they are directed to:
Additional mental steps:
continuously changing the weighting values within a predefined transition time upon a change in context,
determining at least some of the weighting values on a basis of at least one parameter of the current context by way of an assignment table,
wherein at least one operating parameter or operating parameter range is assigned,
a weighting of the at least one machine learning model is determined,
determining the weighting of the at least one of the machine learning models, and
detect hands-off state on a basis of at least the captured steering variable;
Insignificant extra-solution activity
output a hands-off state on a basis of at least the captured steering variable;
Application of machine learning models in a generic fashion to apply the abstract idea in a certain technological environment:
wherein one of the at least two trained machine learning models is a general standard model and at least one other of the at least two trained machine learning models is a machine learning model trained for a specific context,
activating at least one further machine learning model depending on the context, wherein the further machine learning model is trained, and
determining the weighting of at least one of the machine learning models based on the current context by a trained third machine learning model; or
Generally linking the use of a judicial exception to a particular technological environment or field of use
a steering system,
a vehicle, and
a vehicle, comprising a steering system with an apparatus
Claims 1-11 do not pass the patent eligibility test. Accordingly, claims 1-11 are rejected under § 101.
CLAIM REJECTIONS—35 U.S.C. § 103
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 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.
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.
Claims 1 and 4-11 are rejected under § 103 as being unpatentable over Cheng (CN115782892A) in view of Hamzeh et al. (US20240330413A1; “Hamzeh”).
As to claim 1, Cheng discloses a method for detecting a hands-off state at a steering wheel, the method which comprises:
capturing a steering variable at the steering wheel (“In step S110, a first signal from the steering wheel angle sensor, a second signal from the torque sensor, and a third signal from the wheel speed sensor are received.” ¶ 31.);
supplying the steering variable as input data to at least two trained machine learning models, wherein the machine learning models are trained to detect a hands-off state on a basis of at least the captured steering variable and to output corresponding output data (“In step S120, based on the first signal, the second signal and the third signal, a trained model is used to detect whether the driver's hands have left the steering wheel.” ¶ 32. “In one embodiment, the trained model includes multiple models, each of which corresponds to different vehicle operating conditions (e.g., flat/bumpy road surface, high speed/low speed, etc.). Therefore, in this embodiment, step S120 may include: determining the current vehicle operating condition; selecting … a corresponding model combination from the trained models according to the current vehicle operating condition; and using … the corresponding model combination to detect whether the driver's hands have left the steering wheel based on the first signal, the second signal and the third signal.” ¶ n0031.);
determining application of the at least two trained machine learning models on a basis of a current context (See at least ¶ n0031.); and
combining the output data to form and provide a hands-off state (See at least ¶ n0031.).
Cheng fails to explicitly disclose:
determining weighting values on a basis of a current context; and
combining the output data in weighted form to form and provide a hands-off state.
Nevertheless, Hamzeh teaches:
determining weighting values on a basis of a current context (“Included in machine learning system 100 are a plurality of machine learning prediction models 105 a-c and associated weight models 110 a, 110 b and 110 c (110 a-c). Each of prediction models 105 a-c provides a predicted response in the form or predictions 115 a, 115 b, and 115 c (115 a-c) to question 125, while weight models 110 a-c provide a weight 120 a-c for the predicted responses 115 a-c, respectively. The weights 120 a-c are then used to determine a weight average sum 130 from which output prediction 135 is derived.” ¶ 17 and FIG. 1.); and
combining the output data in weighted form to form and provide an output (See at least ¶ 17 and FIG. 1.).
Cheng discloses: a method for detecting a hands-off state at a steering wheel, comprising at least the steps of providing steering variable data to at least two machine learning models, wherein the machine learning models are applied based on a current context; and providing, via combining the output of the at least two machine learning models, a hands-off state. Hamzeh teaches: determining weighting values on a basis of a current context; and combining the output data in weighted form to form and provide an output.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Cheng to include the feature of: determining weighting values on a basis of a current context; and combining the output data in weighted form to form and provide an output, as taught by Hamzeh, with a reasonable expectation of success because this feature is useful “to select the most likely correct answer from a plethora of machine learning model outputs, significantly improving the accuracy of the machine learning results.” (Hamzeh, ¶ 24.)
Independent claim 8 is rejected for at least the same reasons as claim 1 as the claims recite similar subject matter but for minor differences.
As to claim 4, Cheng fails to explicitly disclose: determining at least some of the weighting values on a basis of at least one parameter of the current context by way of an assignment table.
Nevertheless, Hamzeh teaches: determining at least some of the weighting values on a basis of at least one parameter of the current context by way of an assignment table (“Once the weights 412a-n are identified, the optimization dataset 402 [i.e., assignment table] is labeled with weights 412a-n in labelling operation 415 to form labeled optimization dataset 420. Labeled optimization dataset 420 is used train weight models that can predict the weights based on input data, as illustrated in FIG. 5.” Emphasis added; ¶ 65 and FIG. 5.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Cheng to include the feature of: determining at least some of the weighting values on a basis of at least one parameter of the current context by way of an assignment table, as taught by Hamzeh, with a reasonable expectation of success because this feature is useful “to select the most likely correct answer from a plethora of machine learning model outputs, significantly improving the accuracy of the machine learning results.” (Hamzeh, ¶ 24.)
As to claim 5, Cheng fails to explicitly disclose: wherein at least one operating parameter or operating parameter range is assigned to at least one of the machine learning models, and a weighting of the at least one machine learning model is determined by taking into account a distance of the operating parameter or operating parameter range from a corresponding parameter of the current context.
Nevertheless, Hamzeh teaches: wherein at least one operating parameter is assigned to at least one of the machine learning models, and a weighting of the at least one machine learning model is determined by taking into account a distance of the operating parameter from a corresponding parameter of the current context (“During training, the weight models 525a-n may be used to evaluate how the prediction models 305a-n are performing. If the weights for a certain prediction model and one or all its classes are below a certain threshold … of acceptance, those prediction models may be dropped from the total of the prediction model view. In other words, the model may be omitted from system 100 or system 200 of FIGS. 1 and 2, respectively. The omission of a prediction model may be implemented by simply setting the weights provided by the prediction model's weight model to always be zero for a particular class.” Emphasis added; ¶ 69. Note: The threshold of acceptance represents a distance of an operating parameter (i.e., acceptability) from a corresponding parameter (i.e., a threshold of acceptability) given a certain context.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Cheng to include the feature of: wherein at least one operating parameter is assigned to at least one of the machine learning models, and a weighting of the at least one machine learning model is determined by taking into account a distance of the operating parameter from a corresponding parameter of the current context , as taught by Hamzeh, with a reasonable expectation of success because this feature is useful “to select the most likely correct answer from a plethora of machine learning model outputs, significantly improving the accuracy of the machine learning results.” (Hamzeh, ¶ 24.)
As to claim 6, Cheng discloses: activating at least one further machine learning model depending on the context, wherein the further machine learning model is trained to detect and output a hands-off state on a basis of at least the captured steering variable, and deactivating the same (“In one embodiment, the trained model includes multiple models, each of which corresponds to different vehicle operating conditions (e.g., flat/bumpy road surface, high speed/low speed, etc.). Therefore, in this embodiment, step S120 may include: determining the current vehicle operating condition; selecting … a corresponding model combination from the trained models according to the current vehicle operating condition; and using … the corresponding model combination to detect whether the driver's hands have left the steering wheel based on the first signal, the second signal and the third signal.” ¶ n0031. Note: Necessarily, the model or combination of models which do not correspond to the current vehicle operating condition are deactivated as they are not used. When the combination of models do correspond to the current vehicle operating condition, then they are activated.).
As to claim 7, Cheng fails to explicitly disclose: determining the weighting of at least one of the machine learning models based on the current context by a trained third machine learning model.
Nevertheless, Hamzeh teaches: determining a weighting of at least one of the machine learning models based on a current context by a trained third machine learning model (“As illustrated in FIG. 5 , labeled optimization dataset 420 is used to train weight models 525 a, 525 b, 525 c, 525 d, 525 e, . . . 525 n (525 a-n) to predict the weight that should be provided in order to maximize the accuracy of the prediction from the group of prediction models 305 a-n. … [A]s described above, weight models 110 a-c each provide a weight vectors W that includes three weights, one weight for each of the “close case,” “escalate case” and “on hold” classes of responses.” ¶ 66 and FIG. 5.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Cheng to include the feature of: wherein at least one operating parameter is assigned to at least one of the machine learning models, and a weighting of the at least one machine learning model is determined by taking into account a distance of the operating parameter from a corresponding parameter of the current context , as taught by Hamzeh, with a reasonable expectation of success because this feature is useful “to select the most likely correct answer from a plethora of machine learning model outputs, significantly improving the accuracy of the machine learning results.” (Hamzeh, ¶ 24.)
As to claim 9, Cheng discloses a steering system, comprising an apparatus according to claim 8 (“The output signal of the steering wheel angle sensor can be applied to multiple systems in the vehicle, including ESP stability control, EPS/EHPS, active steering, four-wheel steering, automatic parking and other steering-related systems.” ¶ n0026. Note: The successful operation of Cheng requires that the above-cited steering system is part of Cheng’s overall system – and therefore comprises the claimed apparatus – as otherwise Cheng would be unable to obtain steering wheel data from said steering system.).
As to claim 10, Cheng discloses a vehicle, comprising an apparatus according to claim 8 (“The final trained model is obtained and deployed in the vehicle ECU.” ¶ n0030. Note: The successful operation of Cheng requires that the above-cited vehicle (necessarily comprising a vehicle ECU) is part of Cheng’s overall system – and therefore comprises the claimed apparatus – as otherwise Cheng would be inoperable as it pertains to an innately vehicular system.).
As to claim 11, Cheng discloses a vehicle, comprising a steering system with an apparatus according to claim 8 (“Referring to Figure 2, Figure 2 shows a structural schematic diagram of a steering wheel hands-off detection device 2000 according to an embodiment of the present invention. As shown in Figure 2, the steering wheel hands-off detection device 2000 includes: a first receiving device 210 and a detection device 220, wherein the first receiving device 210 receives a first signal from a steering wheel angle sensor, a second signal from a torque sensor, and a third signal from a wheel speed sensor; and the detection device 220 is used to detect whether the driver's hands have left the steering wheel based on the first signal, the second signal, and the third signal, using a trained model.” ¶ n0036. See also ¶¶ n0026 and n0030 Note: The successful operation of Cheng requires that the above-cited vehicle (necessarily comprising a vehicle ECU) and steering system are part of Cheng’s overall system – and therefore comprise the claimed apparatus – as otherwise Cheng would be inoperable as it pertains to an innately vehicular system which innately interacts with the steering system.).
Claim 2 is rejected under § 103 as being unpatentable over Cheng in view of Hamzeh as applied to claim 1 – further in view of Hong et al. (US20200089653A1; “Hong”)
As to claim 2, the combination of Cheng and Hamzeh fails to explicitly disclose: continuously changing the weighting values within a predefined transition time upon a change of the context.
Nevertheless, Hong teaches: continuously changing weighting values within a predefined transition time upon a change in context (“An AI unit may obtain weight values applied to result values output from a plurality of models (for example, first to third models) 310, 320, and 330, based on sensing information input to the plurality of models 310, 320, and 330.” ¶ 138. “For example, when the first sensing information 910 is data obtained by sensing the first class S1, the loss of each of the first sensing information input to the auto encoder and the result value output from the auto encoder may be small. Also, when the loss is small, an uncertainty of the result value of the first model may be a low level.” ¶ 144. “Moreover, when the uncertainty of the result value of the first model is a low level, the AI unit may output a weight value corresponding to a high level.” ¶ 145. See also ¶¶ 137, 139-143, 146-150 and FIG. 9. Note: The sensing information provide context against which weighting values are changed. That is, when the sensing information changes (e.g., from high loss to low loss, etc.), then the weighting values are continuously changed in turn. Furthermore, the successful operation of Hong’s invention requires that the changing of weighting values upon a change in context is performed in a continuous manner within a transition time as otherwise the invention would be inoperable to account for a change in context.)
Cheng discloses: a method for detecting a hands-off state at a steering wheel, comprising at least the steps of providing steering variable data to at least two machine learning models, wherein the machine learning models are applied based on a current context; and providing, via combining the output of the at least two machine learning models, a hands-off state. Hamzeh teaches: determining weighting values on a basis of a current context; and combining the output data in weighted form to form and provide an output. Hong teaches: continuously changing weighting values within a predefined transition time upon a change in context.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Cheng and Hamzeh to include the feature of: continuously changing weighting values within a predefined transition time upon a change in context, as taught by Hong, with a reasonable expectation of success because this feature is useful for “providing an artificial intelligence (AI) device which controls weight values of result values of a plurality of models in an ensemble model which combines the result values of the plurality of models to output a final result value,” thereby reducing uncertainty of the final result value. (Hong, ¶ 11; see also ¶ 10.).
Claim 3 is rejected under § 103 as being unpatentable over Cheng in view of Hamzeh as applied to claim 1 – further in view of Zhao et al. (US20200380354A1; “Zhao”).
As to claim 3, the combination of Cheng and Hamzeh fails to explicitly disclose: wherein one of the at least two trained machine learning models is a general standard model and at least one other of the at least two trained machine learning models is a machine learning model trained for a specific context.
Nevertheless, Zhao teaches: wherein one of the at least two trained machine learning models is a general standard model and at least one other of the at least two trained machine learning models is a machine learning model trained for a specific context (“A computer-implemented method for detecting an operation tendency is disclosed. The method includes preparing a general model for generating a general anomaly score. The method also includes preparing a specific model, for generating a specific anomaly score, trained with a set of a plurality of operation data related to operation by a target operator.” Abstract; see also FIG. 1.).
Cheng discloses: a method for detecting a hands-off state at a steering wheel, comprising at least the steps of providing steering variable data to at least two machine learning models, wherein the machine learning models are applied based on a current context; and providing, via combining the output of the at least two machine learning models, a hands-off state. Hamzeh teaches: determining weighting values on a basis of a current context; and combining the output data in weighted form to form and provide an output. Zhao teaches: providing a general standard model and a machine learning model trained for a specific context.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Cheng and Hamzeh to include the feature of: wherein one of the at least two trained machine learning models is a general standard model and at least one other of the at least two trained machine learning models is a machine learning model trained for a specific context, as taught by Zhao, with a reasonable expectation of success because this feature is useful for improving the accuracy of an output of an ensemble of machine learning models. (See Zhao, ¶¶ 139 and 141.)
CONCLUSION
Any inquiry concerning this communication or earlier communications from the Examiner should be directed to Mario C. Gonzalez whose telephone number is (571) 272-5633. The Examiner can normally be reached M–F, 10:00–6:00 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, Fadey S. Jabr, can be reached on (571) 272-1516. 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.
/MARIO C GONZALEZ/Examiner, Art Unit 3668
1 Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision) (precedential)