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 .
Response to Arguments
Applicant’s arguments with respect to claims 1-7 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Response to Amendment
The amendment filed 6/6/2026 is objected to under 35 U.S.C. 132(a) because it introduces new matter into the disclosure. 35 U.S.C. 132(a) states that no amendment shall introduce new matter into the disclosure of the invention. The added material which is not supported by the original disclosure is as follows: “processing circuitry” which is not described or mentioned in the specification. The “function” in the purpose information is never described in the specification.
Applicant is required to cancel the new matter in the reply to this Office Action.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
Claims 1-7 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. In claim 1, Applicant claims “processing circuitry” which is not described or mentioned in the specification. The “function” in the purpose information is never described in the specification. The phrase, “minimally changing” the existing model, is not described in the specification and it is not a term of art. What it means when a performance index is “satisfied” is never described. How to “compare the adaptive autonomous agent requirement information with meta information” is never described in the specification. The term “adaptive training data” is not described in the specification and it is not a term of art. The phrase “divide the artificial intelligence model into a plurality of modules…” is not described and not a term of art.
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.
Claims 1-7 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.
The terms “selectively” and “minimally in claim 1 is a relative term which renders the claim indefinite. The term “selectively” and “minimally” are not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. These terms aren’t terms of art and there is no way to discern what they mean, and what they are being compared to when they are claimed.
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.
Claims 1-6 are rejected under 35 U.S.C. 103 as being unpatentable over US20220114495A1 to Nurvitadhi et al (Nur), US20180364654A1 to Locke et al and WO2022087788A1 (’88).
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over US20220114495A1 to Nurvitadhi et al (Nur), US20180364654A1 to Locke et al, WO2022087788A1 (’88) and US20230334320A1 to Zhang et al.
Nur teaches claim 1. (Currently Amended) A module-based prefabricated artificial intelligence development system comprising:
an adaptive autonomous agent configured to perceive a surrounding environment (Nur fig. 8)
an adaptive artificial intelligence development apparatus including processor circuitry and memory; and (Nur fig. 1-3)
an AI module hub including processor circuitry, an AI topology storage, and an AI module storage, (Nur fig. 1-3)
wherein the memory stores instructions which, when executed by the processor circuitry of the adaptive artificial intelligence development apparatus, cause the adaptive artificial intelligence development apparatus to:
obtain adaptive autonomous agent requirement information for the adaptive autonomous agent, the adaptive autonomous agent requirement information including (i) environment information recognized through the at least (Nur para 153 “802, at which the ML system configuration circuitry 300 receives a request to execute a machine-learning (ML) workload…” Nur para 140 “a request, etc., indicative of a desired AI/ML operation (e.g., a desire to do image processing without specifying the initial AI model).” Nur para 157 “810, the ML system configuration circuitry 300 determines whether the evaluation parameter satisfies a threshold. For example, the configuration evaluation circuitry 340 can determine whether an evaluation parameter, such as an accuracy parameter, has a value that satisfies an evaluation parameter threshold, such as an accuracy threshold (e.g., an accuracy parameter threshold).” The function and performance index are the evaluation parameter and threshold. The environment and state are the information indicative of a desired AI/ML operation.)
cause the AI module hub to compare the adaptive autonomous agent requirement information with meta information stored in the AI module hub, and receive, from the AI module hub, an AI topology and a plurality of AI modules corresponding to the adaptive autonomous agent requirement information; (Nur 154 “At block 804, the ML system configuration circuitry 300 generates a first configuration of one or more ML models based on the ML workload.” Generating “based on the ML workload” is the same as Applicant’s “compar[ing]”. The independent “ML models” in Nur are the topology and AI modules.)
generate a candidate artificial intelligence model by assembling the plurality of AI modules under the AI topology; (Nur 154 “At block 804, the ML system configuration circuitry 300 generates a first configuration of one or more ML models based on the ML workload.”)
train the candidate artificial intelligence model using adaptive training data by selectively training only newly connected modules while minimally changing an existing part of the candidate artificial intelligence model; (Nur para 166 “At block 912, the ML system configuration circuitry 300 determines hyperparameters to train the ML model.” Selectively and minimally don’t mean anything and aren’t described in the specification.)
determine whether a performance index of the candidate artificial intelligence model satisfies a target performance index corresponding to the purpose information; (Nur para 157 “810, the ML system configuration circuitry 300 determines whether the evaluation parameter satisfies a threshold. For example, the configuration evaluation circuitry 340 can determine whether an evaluation parameter, such as an accuracy parameter, has a value that satisfies an evaluation parameter threshold, such as an accuracy threshold (e.g., an accuracy parameter threshold).”)
in response to the performance index not satisfying the target performance index, store the candidate artificial intelligence model together with performance information, request generation of a new candidate artificial intelligence model by assembling another type of artificial intelligence module under the AI topology, and repeat the assembling, the training, and the determining until the target performance index is satisfied or candidate artificial intelligence models for all cases under the AI topology and the plurality of AI modules have been evaluated; and (Nur para 158 “810, the ML system configuration circuitry 300 determines that the evaluation parameter does not satisfy a threshold, then, at block 812, the ML system configuration circuitry 300 updates an ontology database based on the evaluation parameter. For example, the ontology generation circuitry 350 (FIG. 3) can update the ontology database 208 of FIG. 2 based on the evaluation parameters 226, the proposed HW/SW instance 222 that are associated with the evaluation parameters 226, etc., and/or any combination(s) thereof.” Nur fig. 8 shows the request to generate new configs. Nur par 159 “814, the ML system configuration circuitry 300 adjusts the first configuration based on the evaluation parameter. For example, the ML software configuration circuitry 320 can replace the CNN with a different AI/ML model, add another AI/ML model, change a configuration of the CNN, etc.,”)
in response to the target performance index being satisfied, transmit a final artificial intelligence model to the adaptive autonomous agent, (Nur para 161 “determines that the evaluation parameter satisfies a threshold, control proceeds to block 818 to execute the one or more ML models based on the ML models based on the first configuration”)
Nur doesn’t teach a sensor and actuator.
However, Locke teaches an adaptive autonomous agent configured to perceive a surrounding environment … (i) environment information recognized through the at least one sensor, (ii) state information including execution state information of at least one of the at least one sensor or the at least one actuator… (Locke para 52 “ airside system 130 delivers the supply airflow into one or more zones of building 10 (e.g., via supply ducts 112) without using intermediate VAV units 116 or other flow control elements. AHU 106 can include various sensors (e.g., temperature sensors, pressure sensors, etc.) configured to measure attributes of the supply airflow. AHU 106 can receive input from sensors located within AHU 106 and/or within the building zone and can adjust the flow rate, temperature, or other attributes of the supply airflow through AHU 106 to achieve set-point conditions for the building zone.”)
Nur, Locke and the claims are all using autonomous agents. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to use sensors and actuators in the agent, and as part of the request data, because “artificial intelligence methods can also allow the BMS system to determine correlations between the system data and operations that can yield beneficial operating models for each system and create synergistic results.” Locke para 45.
Nur doesn’t teach decomposing the models.
However, ’88 teaches wherein the processor circuitry of the AI module hub is configured to:
receive an artificial intelligence model developed for various purposes; (‘88 fig. 1 step 110 “the neural network model parser can parse the input neural network model…” Input shows reception.)
extract topology information of the artificial intelligence model and store structural information about the artificial intelligence model in the AI topology storage; (‘88 fig. 1 step 110 “the neural network model parser can parse the input neural network model, such as analyzing the grammatical structure or syntax of the input neural network model to generate model information;” Analysis is extraction of information.)
divide the artificial intelligence model into a plurality of modules; (‘88 fig. 1 step 120 “s120, the computation graph of the neural network is divided into several subgraphs.”)
generate, for each of the plurality of modules, meta information including related topology information, input/output information, position information, and characteristic information; and (‘88 fig. 1 step 120 “The computational graph can be divided into several subgraphs using the adjacency matrix-based clustering graph cut algorithm. As shown in Figure 3, a clustering algorithm is applied on the adjacency matrix to generate clustering results, thereby determining the boundaries of the sub-graphs to be divided.”
store the plurality of modules in association with the meta information in at least one of the AI topology storage or the AI module storage. (’88 summary of invention “; the topological feature vector of the at least one divided subgraph and its optimization strategy are associated and added to in the topology feature library.”)
Nur, ’88 and the claims are all directed to neural networks. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to use subgraphs and store them because “There are often many identical structures in neural network models, and repeated compilation of these identical structures will seriously reduce the efficiency of some online compilation or compilation of very large models.” ’88 background.
Nur teaches claim 2. (Currently Amended) The system of claim 1, wherein the AI topology is received from the AI topology storage and the plurality of AI modules are received from the AI module storage based on comparison between the adaptive autonomous agent requirement information and meta information stored in the AI topology storage and the AI module storage. (Nur para 154 “At block 804, the ML system configuration circuitry 300 generates a first configuration of one or more ML models based on the ML workload. For example, the ML software configuration circuitry 320 (FIG. 3) can identify an AI/ML model such as a CNN from the software search space 218.” The model is received from the search space.)
Nur teaches claim 3. (Currently Amended) The system of claim 1, wherein the AI module hub further includes an artificial intelligence module management unit configured to perform at least one of storage, retrieval, update, or deletion of the AI topology and the plurality of AI modules, and to search for and extract the AI topology and the plurality of AI modules corresponding to the adaptive autonomous agent requirement information. (Nur para 154 “At block 804, the ML system configuration circuitry 300 generates a first configuration of one or more ML models based on the ML workload. For example, the ML software configuration circuitry 320 (FIG. 3) can identify an AI/ML model such as a CNN from the software search space 218.” The model is received from the search space.)
Nur teaches claim 4. (Currently Amended) The system of claim 1, wherein the meta information generated for each of the plurality of modules further includes a description of the corresponding module. (Nur para 154 “At block 804, the ML system configuration circuitry 300 generates a first configuration of one or more ML models based on the ML workload. For example, the ML software configuration circuitry 320 (FIG. 3) can identify an AI/ML model such as a CNN from the software search space 218.”)
Nur teaches claim 5. (Currently Amended) The system of claim 1, wherein the plurality of modules and the meta information are stored together in the AI module storage. (Nur para 154 “For example, the ML software configuration circuitry 320 (FIG. 3) can identify an AI/ML model such as a CNN from the software search space 218.”)
Nur teaches claim 6. (Currently Amended) The system of claim 1, wherein the artificial intelligence model received by the AI module hub is accompanied by model meta information describing the artificial intelligence model, and the model meta information includes at least one of a classification category, a title, or a description text of the artificial intelligence model. (Nur para 154 “For example, the ML software configuration circuitry 320 (FIG. 3) can identify an AI/ML model such as a CNN from the software search space 218.”)
Nur teaches claim 7. (Currently Amended) The system of claim 1, wherein, when the target performance index is not satisfied after candidate artificial intelligence models (Nur para 158 “determines that the evaluation parameter does not satisfy a threshold, then, at block 812, the ML system configuration circuitry 300 updates an ontology database based on the evaluation parameter.” Nur fig 8 shows that eventually a configuration is chosen and deployed.)
Nur doesn’t exhaust all cases or pick the closest to the target performance.
However, Zhang teaches performance index is not satisfied after candidate artificial intelligence models for all cases under the AI topology and the plurality of AI modules have been evaluated, the instructions further cause the adaptive artificial intelligence development apparatus to transmit, to the adaptive autonomous agent, a candidate artificial intelligence model closest to the target performance index from among stored candidate artificial intelligence models. (Zhang fig. 11 and para 99 “In block 1106, the NAS system 102 uses neural architecture search to produce the chosen machine-trained model that satisfies the latency constraint, based on a collection of candidate machine-trained models.” Zhang chooses the best after searching the whole space.)
Zhang, Nur and the claims all search candidate models. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to use the best fit after searching all the models so that an agent isn’t left without a model to control its actions, that way “The… system ultimately selects a neural network architecture that best satisfies specified performance objectives.” Zhang para 1.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Austin Hicks whose telephone number is (571)270-3377. The examiner can normally be reached Monday - Thursday 8-4 PST.
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, Mariela Reyes can be reached at (571) 270-1006. 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.
/AUSTIN HICKS/Primary Examiner, Art Unit 2142
Prosecution Insights