DETAILED ACTION
This Action is a response to the filing received 17 September 2024. Claims 1-15 were originally presented. In a Preliminary Amendment filed 14 October 2024, claims 1-15 were canceled and claims 16-29 were newly added. Claims 16-29 remain pending for examination.
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 .
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. §§ 102 and 103 (or as subject to pre-AIA 35 U.S.C. §§ 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. § 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim 16 is rejected under 35 U.S.C. § 102(a)(1) as being anticipated by H. Koziolek, S. Gruener and V. Ashiwal, "ChatGPT for PLC/DCS Control Logic Generation," 2023 IEEE 28th Int’l Conf. on Emerging Technologies and Factory Automation (ETFA), Sinaia, Romania, 12-15 September 2023, pp. 1-8 (“Koziolek”).
Regarding claim 16, Koziolek teaches: A computer-implemented method for automated generation of a model of computation including a pushdown automaton and/or a state machine and/or a Petri net, the method comprising: generating, via a machine learning model, at least one model of computation based at least on one specification which the at least one model of computation is to fulfill, and based on a prompt (Koziolek, e.g., Abs., “we created 100 LLM prompts in 10 representative categories to analyze control logic generation for of PLCs and DCS from natural language …” See also, e.g., § IV, 4th para., “Process control prompts involve different kinds of analog control, often involving PI or PID controllers. They cover different process control aspects, such as feedforward, cascade and ration control …” See also, e.g., § V, 4th para., “Figure 2 shows an excerpt of the generated code for (‘TRAFFIC CONTROL’). ChatGPT came up with different light colors, despite not being prompted for them. It also implemented a custom state machine for the phases request in the prompt: normal, pedestrian, and emergency …”); and
evaluating the at least one model of computation, resulting in an evaluation result (Koziolek, e.g., § III, 2nd para., “We fed all prompts into an LLM and analyzed the answers for plausibility and correct functionality. We provide a subjective scoring of the answer quality …” See also, e.g., § VI, 7th para., “Syntactical problems can be easily identified by attempting to compile the generated code. Semantic problems may be harder to find. Testing the control code without physical I/O may require simulated input signal values …”).
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 17-18 are rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Hayashi et al., U.S. 2023/0280985 A1 (“Hayashi”).
Regarding claim 17, the rejection of claim 16 is incorporated, and Koziolek further teaches: wherein the at least one model of computation is configured to map a controlling and/or regulation and/or monitoring of a technical system including a cyber-physical system (Koziolek, e.g., Abs., “we created 100 LLM prompts in 10 representative categories to analyze control logic generation for of PLCs and DCS from natural language …” See also, e.g., § IV, 4th para., “Process control prompts involve different kinds of analog control, often involving PI or PID controllers. They cover different process control aspects, such as feedforward, cascade and ration control …”).
Koziolek does not more particularly teach that the cyber-physical system includes at least one computing unit of a vehicle. However, Hayashi does teach: [the cyber-physical system] which includes at least one computing unit of a vehicle1 (Hayashi, e.g., Abs., “provide a program synthesis framework that generates code programs through a multi-turn conversation between a user and a system …” See also, e.g., ¶¶15-17, describing generating code based on a user-provided specification, and ¶28, “networked system suitable for implementing the program synthesis framework … including the user device 210 … server 230 …” and ¶31, “user device may be implemented as an autonomous driving vehicle …” Examiner’s note: a user of an autonomous driving vehicle engaging with the program synthesis system via the vehicle may be generating code for use by the vehicle) for the purpose of permitting a user to generate program functionality with natural language conversational prompts for a variety of environments without technical expertise (Hayashi, e.g., ¶¶15-19, 28-35).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the cyber-physical system includes at least one computing unit of a vehicle because the disclosure of Hayashi shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for program generation via conversational model prompting to provide that the cyber-physical system includes at least one computing unit of a vehicle for the purpose of permitting a user to generate program functionality with natural language conversational prompts for a variety of environments without technical expertise (Hayashi, Id.).
Regarding claim 18, the rejection of claim 17 is incorporated, and Koziolek further teaches: translating the at least one model of computation into an executable code which is configured to control and/or regulate, and/or monitor the technical system (Koziolek, e.g., § V, 4th para., “Figure 2 shows an excerpt of the generated code for (‘TRAFFIC CONTROL’). ChatGPT came up with different light colors, despite not being prompted for them. It also implemented a custom state machine for the phases request in the prompt: normal, pedestrian, and emergency …”).
Claim 19 is rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Singh et al., U.S. 2022/0206785 A1 (“Singh”).
Regarding claim 19, the rejection of claim 16 is incorporated, but Koziolek does not more particularly teach that the generation of the model of computation via the machine learning model is based at least one counterexample that violates the specification. However, Singh does teach: wherein the generation, via the machine learning model, of the at least one model of computation is based at least on at least one counterexample that violates the specification (Singh, e.g., ¶68, “techniques such as symbolic model checking may be employed to determine whether the pre-migration and post-migration versions of the source code file are, from a mathematical standpoint, semantically equivalent (e.g., will generate the same or consistent output given the same input). Various symbolic model checking techniques may be employed, including but not limited to … Counterexample guided abstraction refinement …”) for the purpose of evaluating a plurality of candidate source code outputs from a machine learning model for whether they perform expected program functions (Singh, e.g., ¶¶67-70).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the generation of the model of computation via the machine learning model is based at least one counterexample that violates the specification because the disclosure of Singh shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for using machine learning models to generate program code to provide that the generation of the model of computation via the machine learning model is based at least one counterexample that violates the specification for the purpose of evaluating a plurality of candidate source code outputs from a machine learning model for whether they perform expected program functions (Singh, Id.).
Claims 20 and 24-25 are rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Andre et al., U.S. 12,217,029 B1 (“Andre”).
Regarding claim 20, the rejection of claim 16 is incorporated, but Koziolek does not more particularly teach retaining the model of computation according to a predetermined criterion based on the evaluation result, or optionally discarding the model of computation. However, Andre does teach: retaining the at least one model of computation according to a predetermined criterion based on the evaluation result (Andre, e.g., 1:31-2:16, “processing a natural language input … requests generation of interfacing source code … processing the interface request … based on a first domain model … to generate a first pool of candidate code snippets … generate a second pool of candidate code snippets … generating a plurality of candidate instances of interfacing source code … each candidate software interface comprises a different permutation of candidate code snippets … output conveys one or more of the candidate instances of interfacing source code … method may include selecting, for inclusion in the output, a given candidate software interface … based on one or more criteria …” Examiner’s note: those candidates meeting the criteria are retained); and
otherwise, optionally discarding the at least one model of computation (Andre, e.g., 2:33-41, “the method may further include eliminating one or more of the plurality of candidate instances … from consideration for inclusion in the output …”) for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, e.g., 1:31-3:2).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide for retaining the model of computation according to a predetermined criterion based on the evaluation result, or optionally discarding the model of computation because the disclosure of Andre shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for generating code snippets via a domain model to provide for retaining the model of computation according to a predetermined criterion based on the evaluation result, or optionally discarding the model of computation for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, Id.).
Regarding claim 24, the rejection of claim 16 is incorporated, but Koziolek does not more particularly teach generating a plurality of models of computation and evaluating the plurality of models of computation. However, Andre does teach: wherein the generating of the at least one model of computation includes generating a plurality of models of computation, wherein the evaluating of the at least one model of computation includes evaluating the plurality of models of computation (Andre, e.g., 1:31-2:16, “processing a natural language input … requests generation of interfacing source code … processing the interface request … based on a first domain model … to generate a first pool of candidate code snippets … generate a second pool of candidate code snippets … generating a plurality of candidate instances of interfacing source code … each candidate software interface comprises a different permutation of candidate code snippets … output conveys one or more of the candidate instances of interfacing source code … method may include selecting, for inclusion in the output, a given candidate software interface … based on one or more criteria …” Examiner’s note: the candidate code instances are the output of the model; that the output may be the model of computation is taught by Koziolek as cited in the rejection of claim 16 and incorporated herein) for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, e.g., 1:31-3:2).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide for generating a plurality of models of computation and evaluating the plurality of models of computation because the disclosure of Andre shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for generating code snippets via a domain model to provide for generating a plurality of models of computation and evaluating the plurality of models of computation for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, Id.).
Regarding claim 25, the rejection of claim 24 is incorporated, but Koziolek does not more particularly teach that the evaluation of the plurality of models of computation includes selecting a model of computation as the model of computation from the plurality according to a predetermined criterion. However, Andre does teach: wherein the evaluation of the plurality of models of computation includes: selecting a model of computation from the plurality of models of computation according to a predetermined criterion, wherein the selected model of computation is the at least one generated model of computation (Andre, e.g., 1:31-2:16, “processing a natural language input … requests generation of interfacing source code … processing the interface request … based on a first domain model … to generate a first pool of candidate code snippets … generate a second pool of candidate code snippets … generating a plurality of candidate instances of interfacing source code … each candidate software interface comprises a different permutation of candidate code snippets … output conveys one or more of the candidate instances of interfacing source code … method may include selecting, for inclusion in the output, a given candidate software interface … based on one or more criteria …” Examiner’s note: the candidate code instances are the output of the model; that the output may be the model of computation is taught by Koziolek as cited in the rejection of claim 16 and incorporated herein) for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, e.g., 1:31-3:2).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the evaluation of the plurality of models of computation includes selecting a model of computation as the model of computation from the plurality according to a predetermined criterion because the disclosure of Andre shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for generating code snippets via a domain model to provide that the evaluation of the plurality of models of computation includes selecting a model of computation as the model of computation from the plurality according to a predetermined criterion for the purpose of evaluating a plurality of generated candidate code snippets to determine whether one or more of the candidates meets a threshold for further consideration by a user (Andre, Id.).
Claim 21 is rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Ataei et al., U.S. 2025/0028882 A1 (“Ataei”).
Regarding claim 21, the rejection of claim 16 is incorporated, but Koziolek does not more particularly teach that the evaluation of the model of computation includes checking whether a syntax and/or grammar of the model of computation is correct via a simulator for an automaton. However, Ataei does teach: wherein the evaluation of the at least one model of computation includes: checking whether a syntax and/or grammar of the at least one model of computation is correct via a simulator for an automaton (Ataei, e.g., ¶¶36-38, “Intent engine 122 determines a user intent for a simulation based on one or more natural-language requests … also determine one or more appropriate simulation tasks … simulation tools, such as one or more solvers and/or simulation models … simulation engine 124 first executes a test run of the solver(s) and/or simulation model(s). During the test run, simulation 124 can check for errors in the solver(s) and/or simulation model(s), such as syntax errors …”) for the purpose of evaluating the output of the model for verification of successfully performing the functions as requested by a user’s natural language text input (Ataei, e.g., ¶¶35-41).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the evaluation of the model of computation includes checking whether a syntax and/or grammar of the model of computation is correct via a simulator for an automaton because the disclosure of Ataei shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for generating simulation code using a generative language model and natural language text input to provide that the evaluation of the model of computation includes checking whether a syntax and/or grammar of the model of computation is correct via a simulator for an automaton for the purpose of evaluating the output of the model for verification of successfully performing the functions as requested by a user’s natural language text input (Ataei, Id.).
Claims 22-23 are rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Koenig et al., U.S. 2024/0193071 A1 (“Koenig”).
Regarding claim 22, the rejection of claim 16 is incorporated, but Koziolek does not more particularly teach that the evaluation of the model of computation includes checking whether a formal specification is met by the model of computation via a model checker. However, Koenig does teach: wherein the evaluation of the at least one model of computation includes: checking whether a formal specification is met by the at least one model of computation via a model checker (Koenig, e.g., ¶¶8-12, “provision of a computer-implemented automated driving function in the form of a software component … provision of a model checker representation of the software component that is limited by the environment model, specification of a formal requirement as an input for a model checking method and application of the model checking method to the model checker representation to analyze the software component with respect to the specified formal requirement”) for the purpose of utilizing a model of a software component to identify edge cases, generate test data, and perform verification of whether the software component complies with a formal specification (Koenig, e.g., ¶¶6-19 et seq.).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the evaluation of the model of computation includes checking whether a formal specification is met by the model of computation via a model checker because the disclosure of Koenig shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for using models of software components to generate test data to provide that the evaluation of the model of computation includes checking whether a formal specification is met by the model of computation via a model checker for the purpose of utilizing a model of a software component to identify edge cases, generate test data, and perform verification of whether the software component complies with a formal specification (Koenig, Id.).
Regarding claim 23, the rejection of claim 22 is incorporated, but Koziolek does not more particularly teach outputting a counterexample when the check yields a result that the formal specification is not met by the model of computation. However, Koenig does teach: outputting a counterexample when the check yields a result that the formal specification is not met by the at least one model of computation2 (Koenig, e.g., ¶17, “method … requires a model checking analysis of a computer-implemented driving function, it actually only concerns those cases in which the proof of correctness has failed. In these cases, the model checking method not only provides the information that the program code or the software component does not comply with the specified requirement, but also the corresponding counterexample in the form of states and state transitions of the software component that have contributed to non-compliance …”) for the purpose of utilizing a model of a software component to identify edge cases, generate test data, and perform verification of whether the software component complies with a formal specification (Koenig, e.g., ¶¶6-19 et seq.).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide for outputting a counterexample when the check yields a result that the formal specification is not met by the model of computation because the disclosure of Koenig shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for using models of software components to generate test data to provide for outputting a counterexample when the check yields a result that the formal specification is not met by the model of computation for the purpose of utilizing a model of a software component to identify edge cases, generate test data, and perform verification of whether the software component complies with a formal specification (Koenig, Id.).
Claims 26-27 are rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Gorinski et al., W.O. 2024/213252 A1 (“Gorinski”).
Regarding claim 26, Koziolek teaches: A computer-implemented method for further training a machine learning model, wherein the machine learning model is configured to generate at least one model of computation including a pushdown automaton and/or a state machine and/or a Petri net, at least based on at least one specification that the at least one model of computation is to fulfill, and based on a prompt, the method comprising the following steps: …
wherein the at least one model of computation was generated and evaluated according to a method for automated generation of a model of computation, the method for automated generation of the model of computation including: generating, via the machine learning model, the at least one model of computation based at least on one specification which the at least one model of computation is to fulfill, and based on a prompt (Koziolek, e.g., Abs., “we created 100 LLM prompts in 10 representative categories to analyze control logic generation for of PLCs and DCS from natural language …” See also, e.g., § IV, 4th para., “Process control prompts involve different kinds of analog control, often involving PI or PID controllers. They cover different process control aspects, such as feedforward, cascade and ration control …” See also, e.g., § V, 4th para., “Figure 2 shows an excerpt of the generated code for (‘TRAFFIC CONTROL’). ChatGPT came up with different light colors, despite not being prompted for them. It also implemented a custom state machine for the phases request in the prompt: normal, pedestrian, and emergency …”); and
evaluating the at least one model of computation, resulting in the evaluation result (Koziolek, e.g., § III, 2nd para., “We fed all prompts into an LLM and analyzed the answers for plausibility and correct functionality. We provide a subjective scoring of the answer quality …” See also, e.g., § VI, 7th para., “Syntactical problems can be easily identified by attempting to compile the generated code. Semantic problems may be harder to find. Testing the control code without physical I/O may require simulated input signal values …”).
Koziolek does not more particularly teach adapting the machine learning model based on the model of computation and its evaluation result, the evaluation result from evaluating the model of computation. However, Gorinski does teach: adapting the machine learning model at least based on at least one model of computation and at least one evaluation result, wherein the at least one evaluation result results from evaluating the at least one model of computation (Gorinski, e.g., 2:7-14, “input natural language data to the code synthesis model to form output code corresponding to the natural language data; determine the result of one or more functional tests for the output code or a part thereof; and update the code synthesis model based on the result of the one or more functional tests …” See also, e.g., 2:26-30, “Using reinforcement learning, small changes to the code synthesis model which may impact the output code may be explored, using the result of the functional tests as the reward, or as part of the overall reward function …”) for the purpose of utilizing reinforcement learning to iteratively update a code synthesis model for more effective generation of code outputs based on natural language inputs (Gorinski, e.g., 2:7-3:15).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide for adapting the machine learning model based on the model of computation and its evaluation result, the evaluation result from evaluating the model of computation because the disclosure of Gorinski shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for training and using a model to generate code to provide for adapting the machine learning model based on the model of computation and its evaluation result, the evaluation result from evaluating the model of computation for the purpose of utilizing reinforcement learning to iteratively update a code synthesis model for more effective generation of code outputs based on natural language inputs (Gorinski, Id.).
Regarding claim 27, the rejection of claim 26 is incorporated, but Koziolek does not more particularly teach that the adaptation of the machine learning model includes calculating at least one reward based on the evaluation result, and adapting the machine learning model based on the specification and the reward. However, Gorinski does teach: wherein the adaptation of the machine learning model based at least on the at least one specification and the at least one evaluation result includes: calculating at least one reward based at least on the at least one evaluation result (Gorinski, e.g., 11:19-28, “problem description 302 in natural language form is presented to model 301 as input, for which model 301 produces a solution or a number of solutions 303, in the form of programming language code … assesses whether the generated solution or solutions correctly solve the problem defined by a problem description. The evaluation result may be used in one or more of the following two ways:” See also, e.g., 12:1-4, “evaluation result for the functional tests 306 can be used as part of the reward signal of a reinforcement learning process …”); and
adapting the machine learning model based at least on the at least one specification and the at least one reward (Gorinski, e.g., 12:1-4, “The code synthesis model 301 can therefore be updated based on (i) the result of the functional tests 306 …” Examiner’s note: as discussed above, the evaluation result is an assessment of whether the code correctly solves the problem of the problem description (specification); thus, the reward, based on the evaluation result, updates the model based on the reward and the specification) for the purpose of utilizing reinforcement learning to iteratively update a code synthesis model for more effective generation of code outputs based on natural language inputs (Gorinski, e.g., 2:7-3:15).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide that the adaptation of the machine learning model includes calculating at least one reward based on the evaluation result, and adapting the machine learning model based on the specification and the reward because the disclosure of Gorinski shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for training and using a model to generate code to provide that the adaptation of the machine learning model includes calculating at least one reward based on the evaluation result, and adapting the machine learning model based on the specification and the reward for the purpose of utilizing reinforcement learning to iteratively update a code synthesis model for more effective generation of code outputs based on natural language inputs (Gorinski, Id.).
Claims 28-29 are rejected under 35 U.S.C. § 103 as being unpatentable over Koziolek in view of Mennen et al., W.O. 2025/046110 A1 (“Mennen”).
Claims 28-29 are rejected for the reasons given in the rejection of claim 16 above. Examiner notes that while Koziolek clearly discloses a computer-implemented method (i.e., the use of machine learning models via promptings in order to generate output for evaluation, the output including models of computation and code), Koziolek does not explicitly set forth the details of a particular computing device or environment. However, with respect to claim 28, Mennen teaches: A computer system configured for automated generate of a model of computation, in particular a pushdown automaton, a state machine, and/or a Petri net (Mennen, e.g., ¶167, “‘components’ of the platform generally refer to functionality implemented within the platform 100 … such as … the code synthesis pipeline … Such components are typically implemented in software, which is to say computer-readable instructions executed on one or more hardware processors. A computer system refers to a computer or set of networked computers, where the/each computer comprises one or more hardware processors … Computer-readable instructions may be stored in non-transitory media …” Examiner’s note: Mennen discloses generating a model of a software program from input requirements, the model of the program being used to generate code. That the model is more specifically, e.g., a state machine is disclosed in Koziolek as discussed in the rejection of claim 16, incorporated herein), the computer system configured to: [[[perform the method of claim 16]]]; and with respect to claim 29, Mennen teaches: A non-transitory computer-readable medium on which is stored a computer program for automated generation of a model of computation including a pushdown automaton and/or a state machine and/or a Petri net, the computer program, when executed by a computer (Mennen, e.g., ¶167, “‘components’ of the platform generally refer to functionality implemented within the platform 100 … such as … the code synthesis pipeline … Such components are typically implemented in software, which is to say computer-readable instructions executed on one or more hardware processors. A computer system refers to a computer or set of networked computers, where the/each computer comprises one or more hardware processors … Computer-readable instructions may be stored in non-transitory media …” Examiner’s note: Mennen discloses generating a model of a software program from input requirements, the model of the program being used to generate code. That the model is more specifically, e.g., a state machine is disclosed in Koziolek as discussed in the rejection of claim 16, incorporated herein), causing the computer to perform the following steps: [[[the steps of the method of claim 16]]] for the purpose of providing a computing platform on which a process for generating source code using generative AI may execute (Mennen, e.g., ¶¶167-171 et seq.).
Therefore, 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 system and method for utilizing prompts to generative AI models to generate model and code output for programmable logic as taught by Koziolek to provide for the details of a particular computing device or environment for using generative AI to output code because the disclosure of Mennen shows that it was known to those of ordinary skill in the pertinent art to improve a system and method for generating program code using generative AI to provide for the details of a particular computing device or environment for using generative AI to output code for the purpose of providing a computing platform on which a process for generating source code using generative AI may execute (Mennen, Id.).
Conclusion
Examiner has identified particular references contained in the prior art of record within the body of this action for the convenience of Applicant. Although the citations made are representative of the teachings in the art and are applied to the specific limitations within the enumerated claims, the teaching of the cited art as a whole is not limited to the cited passages. Other passages and figures may apply. Applicant, in preparing the response, should consider fully the entire reference as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art and/or disclosed by Examiner.
Examiner respectfully requests that, in response to this Office Action, support be shown for language added to any original claims on amendment and any new claims. That is, indicate support for newly added claim language by specifically pointing to page(s) and line number(s) in the specification and/or drawing figure(s). This will assist Examiner in prosecuting the application.
When responding to this Office Action, Applicant is advised to clearly point out the patentable novelty which he or she thinks the claims present, in view of the state of the art disclosed by the references cited or the objections made. He or she must also show how the amendments avoid such references or objections. See 37 C.F.R. 1.111(c).
Examiner interviews are available via telephone and video conferencing using a USPTO-supplied web-based collaboration tool. Applicant is encouraged to submit an Automated Interview Request (AIR) which may be done via https://www.uspto.gov/patent/uspto-automated-interview-request-air-form, or may contact Examiner directly via the methods below.
Any inquiry concerning this communication or earlier communication from Examiner should be directed to Andrew M. Lyons, whose telephone number is (571) 270-3529, and whose fax number is (571) 270-4529. The examiner can normally be reached Monday to Friday from 10:00 AM to 6:00 PM ET. If attempts to reach Examiner by telephone are unsuccessful, Examiner’s supervisor, Wei Mui, can be reached at (571) 272-3708. Information regarding the status of an application may be obtained from the Patent Center system. For more information about the Patent Center system, see https://www.uspto.gov/patents/apply/patent-center. If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call (800) 786-9199 (in USA or Canada) or (571) 272-1000.
/Andrew M. Lyons/Primary Examiner, Art Unit 2191
1 Koenig et al., U.S. 2024/0193071 A1, describes provision of a computer-implemented automated driving function in the form of a software component; as a software component is the end result of the method of Koziolek, a combination with Koenig shows that it would have been obvious to apply the machine-learning based software model and code generation methods in the control systems area to the autonomous vehicle area
2 Rabinovitch et al., U.S. 2012/0233587 A1 also teaches model checking verification of software applications involving a statement associated with a verification-related property of the application, and applying the statement to a model of the application to yield an indication of either correctness or a counter example
Prosecution Insights