DETAILED ACTION
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 5/4/2026 has been entered.
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 .
Respond to Amendment
This action is in response to the amendment filed on ----5/4/2026 has been entered.
Claim 1, 10, 19 are amended.
Claim rejection under 35 U.S.C. 101 section has been withdrawn in light of the applicant’s remarks and amendment.
Respond to Argument
Applicant’s arguments, see page 10 – 15, filed on 5/4/2026, with respect to the rejection(s) under U.S.C. 103 have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Ding et al., “Safe Trajectory Generation for Complex Urban Environments Using Spatio-temporal Semantic Corridor”.
Applicant stated in page 11 – 12 that “Ding's flow field provides, for each grid cell, a density value (count of vehicles) and a direction (average heading). This is different from a ‘trajectory value function that specifies a respective value for each of a plurality of possible trajectories in a trajectory space’”, “this cost function is applied during path planning to evaluate individual candidate paths one at a time during a dynamic programming search. It is not a ‘trajectory value function that specifies a respective value for each of a plurality of possible trajectories in a trajectory space.’” Examiner notes that specification of the instant application describes: “the cost or value function can specify costs or values associated with the various possible trajectories in the trajectory space, and may be usable to identify an optimal or desirable trajectory” (specification 0044 of the instant application). Based on the presented claim, the “trajectory value-based flow field” is now maps to the FlowMap path generation framework of Ding which, according to Algorithm 2, sec, Fig. 5 & V.C., “produce multiple candidate left-turn paths that match human driving patterns” and uses “two weighted costs, namely, the density cost and the direction cost, which punish the paths that enter low-density areas or do not match the field direction”, i.e., the cost/value for each of path is a function of the density in the area and the field direction. Ding discloses a system that generates a plurality of path/trajectories with a function that calculate costs (values) associated with each of the generated path/trajectories (trajectory space). Thus, Ding fulfilled the claimed limitation of “ the trajectory value-based flow field having a trajectory value function that specifies a respective value for each of a plurality of possible trajectories in a trajectory space for the map segment”.
Applicant further stated in page 14 – 15 that “Ding does not teach providing trajectory planning parameters to a motion planner at a vehicle.” The arguments 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.
For further details, refer to the claim rejection under 35 U.S.C. 103 section.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1 – 6, 8 – 15 and 17 – 20 are rejected under 35 U.S.C. 103 as being unpatentable over Ding et al., (hereinafter Ding1), “FlowMap: Path Generation for Automated Vehicles in Open Space Using Traffic Flow”, in view of Ding et al., (hereinafter Ding2), “Safe Trajectory Generation for Complex Urban Environments Using Spatio-temporal Semantic Corridor”, Fang et al., (hereinafter Fang), “E2DTC: An End to End Deep Trajectory Clustering Framework via Self-Training”.
Claim 1, Ding discloses: A trajectory planning method for a high-definition (HD) mapping platform, comprising:
obtaining multi-agent vehicle trajectory data associated with a map segment of an HD map (sec I & fig. 3 (a), “The traffic flows consist of a number of vehicle trajectories which are produced by an onboard tracking module”, “traffic flows are easy to obtain. They can be obtained by extending RoadMap with an additional traffic flow layer”; the map with additional detail/layer of information, thus HD map; sec. III “It takes sparse semantic point clouds from vehicles (multi-agents) as input and fuses them into a global lightweight semantic map.”);
constructing a trajectory value-based flow field for the map segment based on the multi-agent vehicle trajectory data, the trajectory value-based flow field having a trajectory value function that specifies a respective value for each of a plurality of possible trajectories in a trajectory space for the map segment (refer to the mapping above & sec I & fig. 2, “we present FlowMap (trajectory value-based flow field), a framework for path generation in open space using traffic flow.”; Algorithm 2, sec, Fig. 5 “produce multiple candidate left-turn paths that match human driving patterns”, Sec. V.C, “we have two weighted costs, namely, the density cost and the direction cost, which punish the paths that enter low-density areas or do not match the field direction”; i.e., using FlowMap as a framework to generate possible trajectories (in trajectory space) for map segment (for example fig. 2 and 3, an intersection). The cost/value for each of path is a function of the density in the area and the field direction);
configuring, for the map segment, … a driving policy layer of the HD mapping platform based on the trajectory value-based flow field, wherein the driving policy layer indicates respective values of trajectories in the map segment (refer to the mapping above, Alg. 2, fig. 4 & sec IV. C. “The output of Algorithm 2 consists of paths originating from each entry lane. Due to the discretization of the grid, the path is not smooth enough for control. To this end, we utilize a local path smoothing [28] based on Quadratic Programming (QP) to smooth the path”; As illustrated in fig. 4(c) to (d), the system of Ding further create smooth trajectories/motions (as driving policy layer) that are smooth enough to control/plan for the motion of the vehicle).
Ding do not explicitly teach:
constructing, by a neural network,
trajectory planning parameters
providing the trajectory planning parameters of the driving policy layer to a motion planner at a vehicle.
Ding2, in the same field of endeavor, explicitly teach:
trajectory planning parameters (sec. II.B., “generate a collision-free ‘tube’ around the initial path and formulate the trajectory smoothing problem into a quadratically constrained quadratic programming (QCQP)”, sec. VI. C., “we adopt a quintic (m=5) piecewise Bezier curve as the trajectory parameterization”; Ding teaches smoothing each trajectory using quadratic programming technique by Ding2 [28], Ding2 specifically teaches that the trajectory is parameterize/formulate to fulfill constraints (Ding2, sec. VI). The combination renders obviousness of the claimed limitation);
providing the trajectory planning parameters of the driving policy layer to a motion planner at a vehicle (sec. VIII., “a trajectory optimization formulation which guarantees the safety and feasibility of the output trajectory”; Fig. 3, the motion of Ego vehicle is controlled/planned based on the optimized/smoothed trajectory i.e., the parameterized/formulated trajectory is for the vehicle to travel safely. Thus, is for the trajectory/motion control/planning of the vehicle.).
Ding and Ding2 both teach trajectory optimization/smoothing for vehicle and are analogous. It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention with a reasonable likelihood of success to further include the details of the trajectory formulation/parameterization taught by Ding2’s in the system of Ding to achieve the claimed teaching. One of the ordinary skill in the art would have motivated to make this modification to “find safe an feasible trajectories so that the vehicle precisely follows the behavior plan and navigates smoothly” (Ding 2, VII.B.).
Ding and Ding2 combination does not explicitly teach:
constructing, by a neural network
Fang, in the same field of endeavor, explicitly teach:
constructing, by a neural network (Fang, sec. I, “GPS-enable devices and mobile computing services, massive volumes of trajectory data are collected to capture the mobility of vehicles … Trajectory clustering, an essential and popular trajectory data analytics task to discover similar trajectory groups”, “trajectory via deep learning representation based on neural networks can be used”).
Ding and Ding2 combination and Fang both teach trajectory clustering among multiple vehicles trajectories and are analogous. It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention with a reasonable likelihood of success to further include the neural network approach of Fang’s teaching in the system of Ding and Ding2 combination to achieve the claimed teaching. One of the ordinary skill in the art would have motivated to make this modification to “achieves superior accuracy and efficiency compared with classical clustering methods” (Fang, abs.).
Claim 2, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 1. The combination further teach: wherein constructing the trajectory value-based flow field for the map segment based on the multi-agent vehicle trajectory data includes assessing values of potential trajectories associated with positions within the map segment according to a trajectory value function (refer to the mapping in Claim 1 & Ding sec. V.C., in this case, weighted costs are calculated values for each potential/possible trajectories. Each trajectories are associated with a position/location in the map segment and the cost is a function of the density in the area and the field direction ).
Claim 3, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 1. The combination further teach: generating an obstacle potential field for the map segment based on the trajectory value-based flow field; and updating an implicit obstacle layer of the HD mapping platform based on the obstacle potential field (Ding, sec. IV.B. & fig. 2, “once a significant change in traffic flow fields is observed, which is typically caused by road construction and road re-routing, A map update trigger will be generated by the traffic flow field generation process, which is then sent to RoadMap. We find that the traffic-flow-based update trigger is very useful for detecting the change in road course and can serve as an important complement for lightweight mapping”; i.e., Ding teaches a use case of FlowMap framework for detect possible position/location/field of obstacles in the area/map segment and update the cloud database with newly detected obstacle (implicit obstacle layer)).
Claim 4, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 3. The combination further teach: generating the obstacle potential field for the map segment based on analysis of flow field divergence in the trajectory value-based flow field (refer to the mapping in Claim 3, Ding, sec. IV.B. “significant change is observed”; i.e., the new vehicle trajectory/flow field is significant difference/divergence from the previously collected traffic flow field trajectory).
Claim 5, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 1. The combination further teach: the multi-agent vehicle trajectory data includes, for each of one or more vehicles: ego trajectory data associated with that vehicle; and neighboring agent trajectory data associated with one or more neighboring vehicles (Ding2, sec. V, “we can use the simulated states of other vehicle as the predicted trajectories”, fig. 3 – 4 ,moving obstacles sec. V.B., “pas between the two dynamic obstacles”, i.e., the trajectory planning of autonomous vehicle takes consideration of the trajectories of other moving vehicles/agents in the proximity. Ding teaches dynamic programming based on the space limit of the environment, Ding2 teaches the consideration of the trajectories/movements/dynamics of neighboring objects/vehicles/agents. The combination renders obviousness of the claimed limitation).
The reason for combination is same as Claim 1.
Claim 6, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 1. The combination further teach: training the neural network based on driven trajectories associated with the map segment (Fang, sec. IV., “raw trajectory generated by a moving object is usually represented as a time ordered sequence of sample GPS points, i.e., T = p1 p2 pT . Each point pi T(1 i T) consists of a pair of spatial coordinates (i.e., latitude and longitude) and its observed timestamp. Here, T denotes the length of a trajectory, i.e., the number of sampled points.”; i.e., the training is based on the past trajectories on the map segment).
The reason for combination is same as Claim 1.
Claim 8, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 1. The combination further teach: the multi-agent vehicle trajectory data includes global navigation satellite system (GNSS) navigation data associated with a plurality of driven trajectories in the map segment (Ding2, VII.A., “test environments are annotated from real satellite maps via QGIS”; one of ordinary skilled in the art appreciate that QGIS is open sources geographic information platform incorporating GPS data. Such understanding can be readily available online for example Bogard, “Visualizing GPS Data with QGIS”).
The reason for combination is same as Claim 1.
Claim 9, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 8. The combination further teach: the GNSS navigation data includes global positioning system (GPS) navigation data (refer to the mapping in Claim 8, the GPS data).
Claim 10 – 15 and 17 – 18 are the corresponding apparatus claim of Claim 1 – 6 and 8 – 9. Ding further teaches at least one memory; and at least one processor communicatively coupled with the at least one memory (Ding, sec. III. & Fig. 2, “The framework consists of three major components: a lightweight semantic mapping module with an additional traffic flow management function, a traffic flow field generation module, and an online path planning and smoothing module.”; the framework provide a mobile signal processing and computation function which inherently include data processor and memory to store data). Claims 10 – 15 and 17 – 18 are rejected with same reason.
Claim 19 – 20 are the corresponding non-transitory computer-readable medium claim of Claim 10 – 11. At least algorithm 1 and 2 of Ding are instructions stored in memory for the processor to process data. Claim 19 – 20 are rejected with same reason.
Claim(s) 7 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Ding et al., (hereinafter Ding), “FlowMap: Path Generation for Automated Vehicles in Open Space Using Traffic Flow”, in view of Ding et al., (hereinafter Ding2), “Safe Trajectory Generation for Complex Urban Environments Using Spatio-temporal Semantic Corridor”, Fang et al., (hereinafter Fang), “E2DTC: An End to End Deep Trajectory Clustering Framework via Self-Training” as applied to claim 6 above, and further in view of Lu et a., (hereinafter Lu), “Learning under Concept Drift: A Review”.
Claim 7, Ding, Ding2 and Fang combination renders obviousness of all the limitation of Claim 6. The combination does not explicitly teach: training the neural network based on the driven trajectories associated with the map segment includes identifying anomalous trajectories among the driven trajectories and training the neural network based on the anomalous trajectories.
Lu, in the same field of endeavor, explicitly teach:
training the neural network based on the driven trajectories associated with the map segment includes identifying anomalous trajectories among the driven trajectories and training the neural network based on the anomalous trajectories (Lu, Fig. 2 & sec. 1, “the three major aspects of concept drift: concept drift detection, understanding and adaptation, as shown in Fig. 2”; sec. 2.1, “’concept drift’ as the problem in which
PNG
media_image1.png
27
262
media_image1.png
Greyscale
”; Fig. 13, “new model is trained with latest data to replace the old model when a concept drift is detected.”; Ding and Fang combination teaches to train a neural network model for clustering trajectories and to detect significant change in the trajectory/clustering pattern. Lu teaches when the concept drift (significant change) happens to the model’s input and output, use new data to train a new model. The combination renders obviousness of the limitation.).
Ding (in view of Ding2 and Fang) and Lu both teach the training and inference stage of machine learning and are analogous. It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention with a reasonable likelihood of success to further include the retraining strategy of Lu’s teaching to the system of Ding (in view of Ding2 and Fang) to achieve the claimed teaching. One of the ordinary skill in the art would have motivated to make this modification in order “to provide more reliable data-driven predictions and decision” (Lu, sec. 1).
Claim 16 is the corresponding apparatus claim of Claim 7. The claims is rejected with same reason.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure: Lv et al., US20180341269, which teaches trajectories planning for autonomous vehicle considering multiple moving agents in the proximity and the trajectories are described as a spline function with parameters.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHIEN MING CHOU whose telephone number is (571)272-9354. The examiner can normally be reached Monday- Friday 9 am - 5 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, PATEL HITESH can be reached on 571-270-5442. 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.
/SHIEN MING CHOU/Examiner, Art Unit 3667
/Hitesh Patel/Supervisory Patent Examiner, Art Unit 3667
8/3/26
Prosecution Insights