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 .
This final rejection is in response to the amendment filed on: 05/20/2026.
Claims 1-20 remain rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claim(s) 1, 4-8 and 10-11, 14, 15, and 18-20 remain rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356).
Claim(s) 2, 3, 12, 13, 16, and 17 remain rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356) in view of Engstrom et al (US Patent: 11447142, issued: Sep. 20, 2022, filed: May 16, 2019).
Claim(s) 9 remain rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356) in view of Naghshvar et al (US Application: US 20200150672, published: May 14, 2020, filed: Nov. 13, 2019).
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 remain rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claim 1 – 101 Analysis:
Claim 1 is directed to a system which performs extraction of spatio-temporal features to infer states of one or more agents and predicting future behaviors for the one or more agents to calculate one or more interactivity scores.
101 Analysis Step 2A, Prong One
Claim 1 recites the following limitations (of which bolded limitations constitute a ‘mental process’ that covers performance of the limitations in the human mind using observation, evaluation, judgment and opinion) and the underlined limitations are interpreted as intended-use.
A system for navigation based on internal state inference and interactivity estimation, comprising: a memory storing one or more instructions; and a processor executing one or more of the instructions stored on the memory to perform training a policy for autonomous navigation by: extracting spatio-temporal features from one or more historical observations of two or more agents within a simulation environment including an ego-agent and one or more non-ego-agents; analyzing the spatio-temporal features to infer one or more internal states of one or more of the agents; as a first set of predicted trajectory distributions predicting one or more future behaviors for one or more of the one or more of the agents in a first scenario including an existence of the ego-agent within the simulation environment based on the spatio-temporal features from the ego-agent; and as a second set of predicted trajectory distributions in a second scenario excluding the existence of the ego- agent within the simulation environment based on spatio-temporal features from one or more of the non-ego-agents and no spatio-temporal features from the ego-agent; calculating one or more interactivity scores for one or more of the agents based on a difference between the first set of predicted trajectory distributions from the second set of predicted trajectory distributions from the second scenario, wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions; controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more of the non-ego-agents, wherein a higher interactivity score between the ego-agent and one or more agents is indicative of a higher probability that the ego-agent will negotiate with the non ego-agent.
More specifically, a person can observe spatio-temporal features and evaluate the features to make a judgement about future behaviors, and then make further evaluations of scenarios and judge interactive scores.
101 Analysis Step 2A, Prong Two
With regards to the additional elements of :
“a memory storing one or more instructions and a processor executing one or more of the instructions stored on the memory to perform training a policy for autonomous navigation …”, these additional elements amount to recitation of a computer to perform the limitations of the method and amount to no more than mere instructions to apply the exception using generic computer component(s) and therefore fails to provide an improvement to the technology or technical field. The courts have identified using the words ‘apply it’ (or an equivalent) with the judicial exception to be insufficient to integrating a judicial exception into a practical application (does not integrate a judicial exception into a practical application).
“controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more of the non-ego-agents”, these additional elements also amount to no more than mere instructions to apply the exception where the ‘ego’ agent could be a generic computer component that is used to apply an instructions to execute ‘control’ instructions (such as control-navigation instructions). It is noted that the ‘controlling’ in the claim does not limit or describe what aspect of the ego-agent is being controlled, and thus it can be interpreted that the ego-agent is directed to execute a type of instruction (such as a ‘control’ type instruction or a control-navigation type instruction). The courts have identified using the words ‘apply it’ (or an equivalent) with the judicial exception to be insufficient to integrating a judicial exception into a practical application (does not integrate a judicial exception in to a practical application).
Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application, and the claim is directed to the judicial exception.
101 Analysis Step 2B
As explained with respect above in Step 2A, Prong Two, there are the additional elements of:
“a memory storing one or more instructions and a processor executing one or more of the instructions stored on the memory to perform training a policy for autonomous navigation …”, and “controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more of the non-ego-agents”, and these additional elements were explained in Step 2A, Prong two to be merely ‘apply it’ (or an equivalent) with the judicial exception using generic computer/generic computer components. The courts have found these types of limitations to be insufficient to qualify as ‘significantly more’ when recited in a claim with a judicial exception (see Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984)
Thus, the additional elements are not considered significantly more than the recited exception and also do not provide an inventive concept.
101 Analysis of claims 2-10
Dependent claims 2-10 do not recite any further limitations that cause the claim(s) to be patent eligible. Rather, the limitations of dependent claims are directed toward additional aspects of the judicial exception that do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception. Therefore, dependent claims 2-10 are not patent eligible under the same rationale as claim 1.
101 Analysis of claim 11
Claim 11 is rejected under similar rationale as claim 1 (as it is an independent claim that is broader than claim 1 without generic computer components).
101 Analysis of claims 12-14
Dependent claims 12-14 do not recite any further limitations that cause the claim(s) to be patent eligible. Rather, the limitations of dependent claims are directed toward additional aspects of the judicial exception that do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception. Therefore, dependent claims 12-14 are not patent eligible under the same rationale as claim 11.
101 Analysis of claim 15
Claim 15 is rejected under similar rationale as claim 1. It is noted that claim 15 additionally recites ‘a controller controlling the of the autonomous vehicle navigation based on internal state inference and interactivity estimation autonomous vehicle according to the policy for autonomous navigation and inputs from a vehicle sensor”.
101 Analysis Step 2A- Prong I: Here, the bolded items above constitute a ‘mental process’ that covers performance of the limitations in the human mind using observation, evaluation, judgment and opinion (a person can make a judgement about a navigation decision based upon evaluation of state inference and interactivity estimation, according to policy data and input sensor data. It is noted, the recitation of ‘controlling the navigation’, when broadly interpreted, can be interpreted as a mental process of making a judgement of a control action/decision, and should the applicant have intended the control to have manipulated specific navigation entities/navigation components of the autonomous vehicle that alter/change navigational movement of the autonomous vehicle, then the examiner suggests the applicant consider making this type of clarification (in the interest of distinguishing from a mental process).
101 Analysis Step 2A- Prong II:
With regards to the underlined, the additional elements of:
‘controller controlling the navigation of the autonomous vehicle’, amounts to recitation of a computer/controller to perform the limitations of the method and amount to no more than mere instructions to apply the exception using generic computer component(s) and therefore fails to provide an improvement to the technology or technical field. It is noted that the ‘controlling’ in the claim does not limit or describe what aspect of the autonomous vehicle is being controlled, and thus it can be interpreted that the controller is directed to execute a type of instruction (such as a ‘control’ type instruction or a control-navigation type instruction). The courts have identified using the words ‘apply it’ (or an equivalent) with the judicial exception to be insufficient to integrating a judicial exception into a practical application (does not integrate a judicial exception in to a practical application).
‘input from a vehicle sensor’, amounts to mere data gathering (i.e: Obtaining information about transactions using the Internet to verify credit card transactions, CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)), which is a form of insignificant extra solution activity. The courts have identified adding insignificant extra solution activit(ies) to the judicial exception to be insufficient to integrating a judicial exception into a practical application (does not integrate a judicial exception in to a practical application).
Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application, and the claim is directed to the judicial exception.
101 Analysis Step 2B
As explained with respect above in Step 2A, Prong Two, there are the additional elements of:
“…controller controlling the navigation of the autonomous vehicle”, this additional element was explained in Step 2A, Prong two to be merely ‘apply it’ (or an equivalent) with the judicial exception using generic computer/generic computer component(s). The courts have found these types of limitations to be insufficient to qualify as ‘significantly more’ when recited in a claim with a judicial exception (see Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984)
“… inputs from a vehicle sensor …”, these additional elements were explained in 2A, Prong two to be insignificant extra solution activity and the courts have found these types of limitations to be insufficient to qualify as ‘significantly more’ when recited in a claim with a judicial exception. (e.g., mere data gathering in conjunction with a law of nature or abstract idea such as a step of obtaining information about credit card transactions so that the information can be analyzed by an abstract mental process, as discussed in CyberSource v. Retail Decisions, Inc., 654 F.3d 1366, 1375, 99 USPQ2d 1690, 1694 (Fed. Cir. 2011)).
Thus, the additional elements are not considered significantly more than the recited exception and also do not provide an inventive concept.
101 Analysis of claims 16-20
Dependent claims 16-20 do not recite any further limitations that cause the claim(s) to be patent eligible. Rather, the limitations of dependent claims are directed toward additional aspects of the judicial exception that do not integrate the judici9al exception into a practical application nor amount to significantly more than the judicial exception. Therefore, dependent claims 16-20 are not patent eligible under the same rationale as claim 15.
Claim Rejections - 35 USC § 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 1-20 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.
With regards to claim 1, the claim requires “negotiate”, which is subjective. For example the term negotiate could be interpreted to mean taking into consideration factors/parameters to make a decision, or the term negotiate could possibly mean a communications has occurred between parties/entities but it is unclear whether the applicant intended that two way bidirectional communication has occurred or even if single communications has occurred. It is also unclear if the applicant intended an agreement has occurred between parties for ‘negotiation’, as the claim only recites an action of negotiation without reciting any terms/conditions for what is hypothetically being agreed upon (in order to make a decision). Lastly, the specification does not define the term ‘negotiate’, and thus, for purposes of examination, the examiner will interpret ‘negotiate’ to encompass: taking into consideration factors/parameters to make a decision/action.
With regards to claims 2-10, they do not resolve the deficiencies of claim 1, and thus they are rejected under similar rationale as claim 1.
With regards to claim 11, it is rejected under similar rationale as claim 1.
With regards to claims 12-14, they do not resolve the deficiencies of claim 11, and thus they are rejected under similar rationale as claim 11.
With regards to claim 15, it is rejected under similar rationale as claim 1.
With regards to claims 16-20, they do not resolve the deficiencies of claim 15, and thus they are rejected under similar rationale as claim 15.
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.
Claim(s) 1, 4-8 and 10-11, 14, 15, and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356).
With regards to claim 1. Hu et al teaches a system for navigation based on internal state inference and interactivity estimation, comprising:
to perform training a policy for autonomous navigation by:
extracting spatio-temporal features from one or more historical observations of two or more agents within a simulation environment including an ego-agent and one or more non-ego agents (page 6321, left column and 6322: observation values are feature values (associated with spatial positioning and trajectory) for a plurality of agents/actors in an environment. Some actors include an ego agent (such as vehicle) and non-ego agent(s) (such as pedestrians));
analyzing the spatio-temporal features to infer one or more internal states of one or more of the agents (page 6322, left column: internal hidden states for one or more actors (agents) are inferred (q(t));
predicting one or more future behaviors for one or more of the agents: … in a first scenario including an existence of the ego-agent within the simulation environment … (page 6322, left column, page 6323, left column: an interactive scenario for more an ego agent such as a vehicle along with another agent (pedestrian) is processed to determine future trajector(ies)/behaviors) and … in a second scenario excluding the existence of the ego-agent within the simulation environment (page 6322, left column: an individual scenario without existence of another actor (agent) is processed to help predict future trajectory(ies)/behaviors);
However Hu et al does not expressly teach … a memory storing one or more instructions; and a processor executing one or more of the instructions stored on the memory …; as a first set of predicted trajectory distributions … based on the spatio-temporal features from the ego-agent; and as a second set of predicted trajectory distributions … based on spatio-temporal features from one or more of the non-ego agents and no spatio-temporal features from the ego-agent; and calculating one or more interactivity scores for one or more of the agents based on a difference between the first set of predicted trajectory distributions from the first scenario and the second set of predicted trajectory distributions from the second scenario, wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions; and controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more of the non-ego-agents, wherein a higher interactivity score between the ego-agent and one or more of the agents is indicative of a higher probability that the ego-agent will negotiate with the non-ego-agent.
Yet Mahjourian et al teaches teach … a memory storing one or more instructions; and a processor executing one or more of the instructions stored on the memory … (paragraphs 0108 and 0110: processor, memory is implemented);
as a first set of predicted trajectory distributions … based on the spatio-temporal features from the ego-agent (paragraphs 0010, 0011, 0034: a first set of distribution (conditional distribution) of trajectory based data (which can include data based upon past states and velocity(ies)) from the ego/query agent and non -ego/non-query agent(s) are generated ); and as a second set of predicted trajectory distributions … based on spatio-temporal features from one or more of the non-ego agents and no spatio-temporal features from the ego-agent (paragraphs 0011, 0012 0034: a second set distribution (marginal) of trajectory based data (that excludes /not-conditioned on the ego/query agent) is generated ); and calculating one or more interactivity scores for one or more of the agents based on a difference between the first set of predicted trajectory distributions from the first scenario and the second set of predicted trajectory distributions from the second scenario (paragraph 0058-0060: interactive score(s) are calculated based on a difference (amount of divergence) between the first and second distributions); and controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more of the non-ego-agents (paragraphs 0015 0016, 0030 and 0031, 0057: ego/query agent navigation behavior is controlled based upon models/predictions that are based on score that help guide the ego/query agent), wherein a higher interactivity score between the ego-agent and one or more of the agents is indicative of a higher probability that the ego-agent will negotiate with the non-ego-agent (paragraph 0053-0057, 092: interactivity scores affect operation of planning a trajectory of a query agent (the claimed ego-agent). More specifically, as also explained in paragraph 0057 and claim 14 of Mahjourian, the query agent (ego-vehicle/autonomous vehicle) is controlled based on the ‘trajectory prediction outputs’ (which are the ‘interactivity scores’) and also the query agent’s resources that control the query agent are affected for non-ego agents above a threshold, such that the quality of the query agent’s final trajectory is improved (paragraph 0056). Citation from paragraph 0057 is provided here for convenience: “the conditional trajectory prediction outputs (e.g., the interactivity scores 170) can be used to guide the control of the query agent … the query agent 112A is an autonomous vehicle, the system 100 can provide (i) the conditional trajectory prediction outputs for the one or more target agents, (ii) data derived from the conditional trajectory prediction outputs for the one or more target agents, or (iii) both to an onboard system of the autonomous vehicle for use in controlling the autonomous vehicle”).
It would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have modified Hu et al’s ability to analyze spatio temporal features to infer state data of one or more agents and to predict future trajectory behavior(s) for the one or more agents in a first and second scenario, as taught by Mahjourian et al. The combination would have allowed Hu et al to have objectively assessed trajectory data between an ego/query agent and target agents in order to help anticipate driver interactions and improved quality of future trajectories for the ego agent (Mahjourian et al, paragraphs 0016 and 0053).
However the combination of Hu et al and Mahjourian et al does not expressly teach “… wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions”.
Yet Kim et al teaches wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions (page 169350, equation # (3): a KL Divergence term within a loss function is used as a weighting value/error of two trajectory distributions to balance the difference magnitude squared term in the loss function).
It would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have modified Hu et al and Mahjourian et al’s ability to determine an interactivity score with KL divergence, such that the score is used as a balancing term (to weight) used in combination with the reconstruction term (difference magnitude squared) within a loss function, as taught by Kim et al. The combination would have allowed construction of new and plausible trajectory datapoints when predicting trajectories.
With regards to claim 4. The system for navigation based on internal state inference and interactivity estimation of claim 1, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein one or more of the historical observations of one or more of the agents is a position or a velocity (as similarly explained in the rejection of claim 1, Mahjourian et al teaches past data can include velocity data ( paragraph 0034)), and is rejected under similar rationale.
With regards to claim 5. The system for navigation based on internal state inference and interactivity estimation of claim 1, Hu et al teaches wherein the extracting the spatio-temporal features from one or more of the historical observations of one or more of the agents is performed by a graph-based encoder (Figure 2, page 6321, section 3: An NMMP module/encoder references an associated set of LSTMs to take observed positional data, and also trajectory of one or more agents to produce encoded actor embedding (NMMP module is an interaction graph having message passing between LSTM embedded trajectory data layers)), and is rejected under similar rationale.
With regards to claim 6. The system for navigation based on internal state inference and interactivity estimation of claim 5, Hu et al teaches wherein the graph-based encoder includes a first long-short term memory (LSTM) layer, a graph message passing layer, and a second LSTM layer (as similarly explained in the rejection of claim 5, NMMP is implemented to use multiple LSTMs and message passing), and is rejected under similar rationale.
With regards to claim 7. The system for navigation based on internal state inference and interactivity estimation of claim 6, Hu et al teaches wherein the graph message passing layer is positioned between the first LSTM layer and the second LSTM layer (as similarly explained in the rejection of claim 5, the NMMP module is an interaction graph having message passing between LSTM embedded trajectory data layers), and is rejected under similar rationale.
With regards to claim 8. The system for navigation based on internal state inference and interactivity estimation of claim 6, Hu et al teaches wherein an output of the first LSTM layer and an output of the second LSTM layer is concatenated to generate final embeddings (Figure 2, page 6321, eq 2a and eq 2b: output of the LSTM layer data is concatenated /accumulated to produce final interacted Actor Embedding data), and is rejected under similar rationale.
With regards to claim 10. The system for navigation based on internal state inference and interactivity estimation of claim 1, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein Kullback-Leibler (KL) divergence is used to measure the difference between the first scenario and the second scenario (as similarly explained in the rejection of claim 1, Mahjourian et al was explained to use Kullback-Leibler to measure difference between two scenarios scenarios), and is rejected under similar rationale.
With regards to claim 11 the combination of Hu et al, Mahjourian et al and Kim et al teaches a computer-implemented method for navigation based on internal state inference and interactivity estimation, comprising training a policy for autonomous navigation by: extracting spatio-temporal features from one or more historical observations of two or more agents within a simulation environment including an ego-agent and one or more non ego-agents; analyzing the spatio-temporal features to infer one or more internal states of one or more of the agents; predicting one or more future behaviors for of the one or more of the agents: as a first set of predicted trajectory distributions in a first scenario including an existence of the ego-agent within the simulation environment based on the spatio-temporal features from the ego-agent; and as a second set of predicted trajectory distributions in a second scenario excluding the existence of the ego-agent within the simulation environment based on spatio-temporal features from one or more of the non-ego agents and no spatio-temporal features from the ego-agent; calculating one or more interactivity scores for one or more of the agents based on a difference between the first set of predicted trajectory distributions from the first scenario and the second set of predicted trajectory distributions from the second scenario, wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions; and controlling the ego-agent based on one or more of the interactivity scores between the ego-agent and one or more non ego-agents, wherein a higher interactivity score between the ego-agent and one or more of the agents is indicative of a higher probability that the ego-agent will negotiate with the non ego-agent, as similarly explained in the rejection of claim 1, and is rejected under similar rationale.
With regards to claim 14. The computer-implemented method for navigation based on internal state inference and interactivity estimation of claim 11, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein one or more of the historical observations of one or more of the agents is a position or a velocity, as similarly explained in the rejection of claim 4, and is rejected under similar rationale.
With regards to claim 15. the combination of Hu et al, Mahjourian et al and Kim et al teaches a navigation based on internal state inference and interactivity estimation autonomous vehicle, comprising: a memory storing one or more instructions; a storage drive storing a policy for autonomous navigation; a processor executing one or more of the instructions stored on the memory to perform autonomous navigation by utilizing the policy for autonomous navigation, wherein the policy for autonomous navigation is trained by: extracting spatio-temporal features from one or more historical observations of two or more agents within a simulation environment including an ego-agent and one or more non-ego agents; analyzing the spatio-temporal features to infer one or more internal states of one or more of the agents; predicting one or more future behaviors for the one or more of the agents: as a first set of predicted trajectory distributions in a first scenario including an existence of the ego-agent within the simulation environment based on the spatio-temporal features from the ego-agent; and as a second set of predicted trajectory distributions in a second scenario excluding the existence of the ego-agent within the simulation environment based on spatio-temporal features from one or more of the non-ego-agents and no spatio-temporal features from the ego-agent; calculating one or more interactivity scores for one or more of the agents based on a difference between the first set of predicted trajectory distributions from the first scenario and the second set of predicted trajectory distributions from the second scenario, wherein the one or more interactivity scores are used as weights of prediction errors for the first and second set of predicted trajectory distributions; and a controller controlling the navigation of the autonomous vehicle based on internal state inference and interactivity estimation autonomous vehicle according to the policy for autonomous navigation and inputs from a vehicle sensor, wherein a higher interactivity score between the ego-agent and one or more of the agents is indicative of a higher probability that the ego-agent will negotiate with the non ego-agent, as similarly explained in the rejection of claim 1, and is rejected under similar rationale.
With regards to claim 18. The navigation based on internal state inference and interactivity estimation autonomous vehicle of claim 15, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein one or more of the historical observations of one or more of the agents is a position or a velocity, as similarly explained in the rejection of claim 4, and is rejected under similar rationale.
With regards to claim 19. The navigation based on internal state inference and interactivity estimation autonomous vehicle of claim 15, the combination of Hu et al and Mahjourian et al teaches wherein the extracting the spatio-temporal features from one or more of the historical observations of one or more of the agents is performed by a graph-based encoder, as similarly explained in the rejection of claim 5, and is rejected under similar rationale.
With regards to claim 20. The navigation based on internal state inference and interactivity estimation autonomous vehicle of claim 19, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein the graph-based encoder includes a first long-short term memory (LSTM) layer, a graph message passing layer, and a second LSTM layer, as similarly explained in the rejection of claim 6, and is rejected under similar rationale.
Claim(s) 2, 3, 12, 13, 16, and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356) in view of Engstrom et al (US Patent: 11447142, issued: Sep. 20, 2022, filed: May 16, 2019).
With regards to claim 2. The system for navigation based on internal state inference and interactivity estimation of claim 1, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein the calculating one or more interactivity scores for one or more of the agents, as similarly explained in the rejection of claim 1, and is rejected under similar rationale.
However Hu et al, Mahjourian et al and Kim et al does not expressly teach … the calculating … is based on counter factual prediction
Yet Engstrom et al teaches … the calculating … is based on counter factual prediction (Abstract, Figure 1, Figure 3, column 1, lines 1-67, column 3, lines 43-53, column 4, lines 38-49, column 12, lines 56-67, column 13, lines 1-10 and column 16, lines 1-32: interaction/behavior between actors are scored (using a computer/processing system) with respect to a counter factual metrics (surprise metric), given sensed/perceived data and the scores are used to determine a decision level of navigation yield-action for the ego-vehicle).
It would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have modified Hu et al, Mahjourian et al and Kim et al’s ability to analyze spatio temporal features to calculate one or more interactivity score(s) and assess how to control the ego vehicle based on the score(s) such that the different sets of the trajectories could have assessed using counter factual methods to glean /arrive-at a desired vehicular /ego-vehicle navigation action), as taught by Engstrom et al. The combination would have allowed for evaluating how surprising a particular action would be to other road users (Engstrom et al, column 3, lines 31-40).
With regards to claim 3. The system for navigation based on internal state inference and interactivity estimation of claim 1, the combination of Hu et al, Mahjourian et al, Kim et al and Engstrom et al teaches wherein one or more of the internal states is an aggressiveness level or a yielding level (as similarly explained in the rejection of claim 2, Engstrom teaches a level of yielding), and is rejected under similar rationale.
With regards to claim 12. The computer-implemented method for navigation based on internal state inference and interactivity estimation of claim 11, the combination of Hu et al, Mahjourian et al, Kim et al and Engstrom et al teaches wherein the calculating one or more interactivity scores for one or more of the agents is based on counter factual prediction, as similarly explained in the rejection of claim 2, and is rejected under similar rationale.
With regards to claim 13. The computer-implemented method for navigation based on internal state inference and interactivity estimation of claim 11, the combination of Hu et al, Mahjourian et al, Kim et al and Engstrom et al teaches wherein one or more of the internal states is an aggressiveness level or a yielding level, as similarly explained in the rejection of claim 3, and is rejected under similar rationale.
With regards to claim 16. The navigation based on internal state inference and interactivity estimation autonomous vehicle of claim 15, the combination of Hu et al, Mahjourian et al, Kim et al and Engstrom et al teaches wherein the calculating one or more interactivity scores for one or more of the agents is based on counter factual prediction, as similarly explained in the rejection of claim 2, and is rejected under similar rationale.
With regards to claim 17. The navigation based on internal state inference and interactivity estimation autonomous vehicle of claim 15, the combination of Hu et al, Mahjourian et al, Kim et al and Engstrom et al teaches wherein one or more of the internal states is an aggressiveness level or a yielding level, as similarly explained in the rejection of claim 3, and is rejected under similar rationale.
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hu et al (“Collaborative Motion Prediction via Neural Motion Message Passing”, publisher: Computer Vision Foundation, published: June 2020, pages 6319-6328) in view of Mahjourian et al (US Application: US 2022/0135086, published: May 5, 2022, filed: Oct. 29, 2021) in view of Kim et al (“Driving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory”, published: Dec. 24, 2021, publisher: IEEE Access, pages 169348-169356) in view of Naghshvar et al (US Application: US 20200150672, published: May 14, 2020, filed: Nov. 13, 2019).
With regards to claim 9. The system for navigation based on internal state inference and interactivity estimation of claim 1, the combination of Hu et al, Mahjourian et al and Kim et al teaches wherein the training the policy for autonomous navigation, as similarly explained in the rejection of claim 1, and is rejected under similar rationale.
However the combination does not expressly teach .. the training … is based on a Partially Observable Markov Decision Process (POMDP).
Yet Naghshvar et al teaches .. the training … is based on a Partially Observable Markov Decision Process (POMDP) (paragraph 0026: training includes modeling as POMDP for driving and other sequential decision making processes).
It would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have modified Hu et al, Mahjourian et al and Kim et al’s ability to implement training for autonomous navigation, such that the training would have been based on POMDP, as taught by Naghshvar et al . The combination would have improved reinforcement learning systems while considering uncertainty of state-action value functions before determining how to proceed with selecting an action corresponding to a state-action value function (Naghshvar et al, paragraph 0005).
Response to Arguments
Applicant's arguments filed 05/20/2026 have been fully considered but they are not persuasive.
With regards to claim 1, and it’s corresponding 35 USC 103 rejection, the applicant argues the combination of Hu, Mahjourian, and Kim failing to teach “"a higher interactivity score between the ego- agent and one or more of the agents is indicative of a higher probability that the ego-agent will negotiate with the non ego-agent", as required by claim 1”. More specifically, the applicant points out that “While Mahjourian does disclose that high interactivity scores represent likelihoods that the behavior of the query agent will affect corresponding agents, there is no reference to any higher probability that an alleged ego agent will act in a specific way, or more specifically, negotiate with an alleged non ego-agent. In fact, Mahjourian is silent with respect to a query agent performing negotiation and thus, cannot possibly disclose this feature of amended claim 1”.
The applicant’s arguments have been considered, however, are not persuasive. First, the examiner respectfully notes that ‘will negotiate’ is an intent of use language, and not an action that is required to be performed, but merely a label for the interactivity score. Secondly, the examiner respectfully points out that the term ‘negotiate’ is subjective , and for purposes of examination the examiner will interpret the ego-agent’s ‘negotiate’ is encompassing an interpretation of an action of an ego-agent that takes-into-account/takes-into-account a variable/obstacle/parameter (i.e non ego agent(s)). Third, at the very least, the ego-agent takes into account one or more non ego agent(s (of high level of interactivity score) when applying a final trajectory for the ego agent (the ego agent’s final trajectory is impacted and enhanced based upon the interactivity score. Citation from paragraph 0057 is provided here for convenience: “the conditional trajectory prediction outputs (e.g., the interactivity scores 170) can be used to guide the control of the query agent … the query agent 112A is an autonomous vehicle, the system 100 can provide (i) the conditional trajectory prediction outputs for the one or more target agents, (ii) data derived from the conditional trajectory prediction outputs for the one or more target agents, or (iii) both to an onboard system of the autonomous vehicle for use in controlling the autonomous vehicle”). Should the applicant have required negotiation to encompass the ego agent passing instructions or receiving instructions based upon the interactivity score in order to affect and control trajectory of both the ego agent and the non ego agent(s), the examiner respectfully suggests the applicant consider clarifying the independent claim(s) should this be the applicant’s intent (please provide a reference to the instant application’s specification for support of this intended feature in a subsequent response as well).
With regards to claim 11 and 15, the applicant argues these claims are allowable for reasons presented by the applicant for claim 1. However this argument is not persuasive since the applicant’s arguments for claim 1 have been shown/explained to be rejected above.
With regards to claims 2-10, 12-14, and 16-20 the applicant argues they are allowable by virtue of their dependency upon one of the independent claims. However this argument is not persuasive since the independent claims have been shown/explained to be rejected above.
With regards to 35 USC 101, the applicant argues the aspect of ‘controlling the ego agent’ and an improvement is implemented due to the newly amended ego agent ‘negotiate with the non-ego agents’, and thus Step 2A, prong two is addressed to show an integration into a practical application /improvement of technological field.
In response, the examiner points out that this argument is not persuasive since as explained above, the claim recites at a high level of generality that the aspect of control (interpreted to encompass applying a computer associated with the ego-agent to execute control logic), and furthermore, applying a computer to a judicial exception (the evaluation of a high score to make a negotiation assessment), is not sufficient to be considered integrating the judicial exception into a practical application.
With regards to Step 2B, the control is recited at a high level of generality and can be considered applying a computer to execute control logic and applying the judicial exception using generic computer/generic computer components. The courts have found these types of limitations to be insufficient to qualify as ‘significantly more’ when recited in a claim with a judicial exception (see Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984)
To reiterate, the applicant appears to be arguing a communications aspect to impact trajectory of ego and non ego agents, however the claim language does not currently require this type of limitation.
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 WILSON W TSUI whose telephone number is (571)272-7596. The examiner can normally be reached Monday - Friday 9 am -6 pm.
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, Adam Queler can be reached at (571) 272-4140. 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.
/WILSON W TSUI/Primary Examiner, Art Unit 2172
Prosecution Insights