DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This is a final office action on the merits. Claims 1-2, 5-14, and 19 are currently pending and are addressed below.
The examiner notes that the fundamentals of the rejection are based on the broadest reasonable interpretation of the claim language. Applicant is kindly invited to consider the reference as a whole. References are to be interpreted as by one of ordinary skill in the art rather than as by a novice. See MPEP 2141. Therefore, the relevant inquiry when interpreting a reference is not what the reference expressly discloses on its face but what the reference would teach or suggest to one of ordinary skill in the art.
Response to Arguments
Applicant’s remarks regarding the rejection of claims 1-20 under 35 U.S.C 101 have been considered but are not persuasive. The examiner has carefully considered the arguments and respectfully disagrees. With respect to the amended claim, applicant contends that a system that simultaneously plans optimal routes for multiple agents in an environment that includes obstacles, while taking into account each agent's past movement history (candidate states in previous time steps) and physical attributes (such as size and speed). While it is challenged that the claims are not directed to a simple route search, the examiner interprets that amended claim 1 continue to recite steps that fall within the mental processes category. For example, “analyzing,” “estimating,” and “selecting” constitute a form of determination, or prediction which constitute mental processes regardless of whether they are applied to “multiple agents,” “obstacles,” or “past time steps” and increasing the type or amount of information analyzed is not enough to remove a claim from an abstract idea.
In addition, applicant asserts that this search cannot be performed mentally or with pen and paper by a human, involving advanced computational processing that handles a vast amount of dynamically changing information (such as the positions, velocities, and attributes of multiple agents, as well as the positions of obstacles) in real time, and predicts future states while considering complex interactions among the agents. The examiner has carefully considered the possibility that the claimed method cannot be performed entirely in the human mind, but would in turn, like to respectfully address whether the operations presented, not particularly the volume of data, or complexity of calculations, correspond to mental acts of observation, evaluation, prediction, and determinative decision-making.
Upon further examining the “generating,” “comparing.” “predicting,” and “selecting” steps recited in amended claim 1, the examiner interprets these are classic mental operations regardless of computational implementation speed. Additionally, the processing of dynamically changing information and complex interactions are implemented using generic computer components, such as a processor, performing ordinary functions of executing software modules and instructions. In applicant’s remarks there is no indication, or limitation that recite and improvement to processor architecture, an improved memory structure, etc. Rather it appears the claim describes the processor and claimed units in functional terms, such as, “configured to generate first characteristics information,” or “configured to search for a route,” and implementing an abstract idea on a general-purpose processor, even with the incorporation of a machine learning model (roadmap constructing model) does not amount to a technological improvement.
Integration of a machine learning model does not grant the claim eligibility unless the claim recites a specific, unconventional training procedure, a non-generic model architecture, improved computational mechanism, or technical problem that is solved. The examiner interprets the “roadmap constructing model” in general functionality that takes in characteristics information and outputs estimated candidate states. There is no indication of how the model furthers structure, training, or improved implementation. As a result of this, the examiner interprets it as merely “apply it” level of performing the abstract analysis.
For the reasons listed above, applicant’s arguments are not persuasive and the rejection of claims 1-20 under 35 U.S.C 101 is maintained.
In light of applicant’s amendment to recite the feature of a “hardware processor” in the respective claims, the claim interpretation of claims 1, 2, 10-12 and 14 under 35 U.S.C 112(f) has been withdrawn.
Applicant’s arguments regarding the rejection of claims 1-20 under 35 U.S.C 103 have been considered but are not persuasive. The examiner has carefully considered applicant’s arguments and respectfully disagrees.
Relative to the argument that states Nemanja and Shai “merely focus on navigation for a single vehicle,” and cannot render multi-agent roadmap construction obvious, it should be noted that the claim language of amended claim 1 is not tied to a specialized multi-agent structure, or planning. Specifically, the claim presents, “handing one agent among the plurality of agents as the target agent,” “designating each of the plurality of agents as the target agent,” and “constructing a roadmap for each agent by repeating the estimation process,” which the examiner interprets as a reuse of the same model for multiple agents, where each “agent” is considered individually. Extending a single-agent routing prediction model to multiple agents by applying the same process repeatedly to each agent is considered to be predictable design choice that is well within the knowledge of a person of ordinary skill in the art. Additionally, there is nothing in the claim that requires coordinated multi-agent actions beyond the iterative reuse of the same prediction functions for the multiple agents.
The examiner has considered the three categories of input identified by the applicant, but interprets the combination of Nemanja and Shai to cover. As a whole, the disclosure of Nemanja and Shai disclose obstacle representations, environmental maps, and occupancy information for planning and prediction. The obstacle inputs are a standard component of their motion-planning systems. In Nemanja, there are various implementations of trajectory planning/prediction using the neural network navigation system which may use computations of time steps (time series), including previous positions (i.e. a single point, multiple points at various distances and/or times of the SDV along the route) and candidate states (see Nemanja, ¶¶ [0036], “prior,” [0064]-[0065], “ahead”). Shai uses state histories, particularly past states as model inputs. Lastly, both references present kinematic factors such as size, velocity, heading, or other targets (aside from the candidate) as inputs. The examiner believes there is motivation to adopt a “dedicated roadmap for each agent,” as the need to forecast trajectory, or plan routes for multiple agents leads to a person of ordinary skill in the art to apply an existing agent route planning model iteratively as it considers the multiple agents.
With respective to applicant’s argument that the invention, “dynamically generates, for each agent, a search space optimized based on interactions upon multiple agents,” the argument is not supported by the claim language, which is being considered under broadest reasonable interpretation. Nothing in claim 1 recites any specific multi-agent interaction modeling, cooperative behavior, collision avoidance between agents, or optimized multi-agent architecture. Instead, the examiner interprets the “first,” “second,” and “third” characteristics information to utilize other agents’ goal and candidate states which serve as additional input features. There is no specific requirement regarding how the inputs are used and does not require any integration beyond providing the data to a model.
As such, the rejection of claims 1-20 under 35 U.S.C 103 is maintained.
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 1-2, 5-14, and 19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
STEP 1
The aforementioned claim 1 is directed toward a system which falls within at least one of the statutory categories.
STEP 2A (PRONG 1)
Claim 1
A route planning system comprising:
an information acquiring unit configured to obtain target information including a start state and a goal state of each of a plurality of agents in a continuous state space
a map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training
a search unit configured to search for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent
wherein the roadmap constructing model includes:
a first processing module configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step
a second processing module configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step
an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated
wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning
wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents
estimating one or more candidate states in a next time step using the roadmap constructing model that has completed training by designating the start state of the one agent represented in the obtained target information as a candidate state of the target agent in an initial target time step
repeating estimating of the candidate state of a next time step using the roadmap constructing model that has completed training by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent
The examiner submits that the foregoing bolded limitations constitute a mental
process because under its broadest reasonable interpretation, the claim covers
performance of the limitations in the human mind. The “constructing,” step encompasses a person mentally mapping a obtained information from the agents within the continuous state space. The recitation of “roadmap constructing model,” is presented at apply it level. For instance, see MPEP 2106.05(f) which provides the following considerations for determining whether a claim simply recites a judicial exception with the words “apply it” (or an equivalent),
such as mere instructions to implement an abstract idea on a computer: (1) whether the
claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of
how a solution to a problem is accomplished; (2) whether the claim invokes computers
or other machinery (in this case augmented reality) merely as a tool to perform an
existing process; and (2) the particularity or generality of the application of the judicial
exception. The “searching,” and “estimating” steps are equivalent to a person looking at the collected data (acquired, obtained, generated, etc.) and forming simple judgements. For instance, in the “searching step,” a person can search for a route by simply observing the data acquired, additionally, the candidate states, or current state of the vehicle, relative to its position, current action, and planned navigation can be observed and interpreted accordingly by a person in time increments. The “repeating” step simply creates an iterative process of the abstract ideas where a person repeated the “searching” and “estimates” until a candidate state reaches a goal state. Thus, claim 1 recites at least one mental process.
STEP 2A (PRONG 2)
Claim 1
A route planning system comprising:
an information acquiring unit configured to obtain target information including a start state and a goal state of each of a plurality of agents in a continuous state space
a map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training
a search unit configured to search for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent
wherein the roadmap constructing model includes:
a first processing module configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step
a second processing module configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step
an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated
wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning
wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents
estimating one or more candidate states in a next time step using the roadmap constructing model that has completed training by designating the start state of the one agent represented in the obtained target information as a candidate state of the target agent in an initial target time step
repeating estimating of the candidate state of a next time step using the roadmap constructing model that has completed training by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent
The examiner submits that the identified additional limitations do not integrate the previously discussed abstract ideas into practical applications. Regarding the additional limitations of, “obtaining,” and “generating,” these are forms of data gathering. Both steps are recited at a high level of generality (i.e., as a general means of receiving, obtaining, or generating acquiring information from generic computer components such as a sensor system) and amount to mere data gathering which is a form of insignificant extra-solution activity. Additionally, the “executing” step is also recited at a high level of generality (i.e., as a general means of transmitting, outputting, or displaying) and encompasses a post solution action, which is also a form of insignificant extra-solution activity. Thus, it is clear that the abstract ideas have not been integrated into practical application.
STEP 2B
Claim 1 does not include additional elements (considered individually and as an ordered combination) that are sufficient to amount to significantly more than the judicial exception for the same reasons to those discussed above. The additional elements such as a processor and recited units (facilities mentioned above) to perform the steps amounts to nothing more than applying the exception using a generic computer component. General application of an exception using a generic computer component cannot provide an inventive concept.
Thus, since claim 1 is: (a) directed towards abstract ideas, (b) does not recite additional elements that integrate the judicial exception into a practical application, and (c) does not recite additional elements that amount to significantly more than the judicial exception, it is clear that claim 1 is directed towards non-statutory subject matter. Regarding independent claims 10-12 and 14 please refer to the rejection above as they are commensurate in scope, with claim 1 directed towards a route planning system, claims 10 and 11 to a route planning method, claim 12 to a roadmap constructing device, and claim 14 to a model generating method.
Dependent claims 2-9, 13, and 15-20 do not recite any further limitations that cause the claims to be patent eligible. The limitations of the dependent claims are directed towards additional aspects of the judicial exception and/or additional elements that do not integrate the judicial exception into a practical application.
As such, claims 1-20 are rejected under 35 U.S.C 101 as being drawn to an abstract idea without significantly more, and thus are ineligible.
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 (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 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 1-2, 5-14, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Djuric Nemanja et al. (US20180107215A1), hereinafter referred to as Nemanja in view of Shalev-Shwartz Shai et al. (US20190291727A1), hereinafter referred to as Shai.
Regarding claim 1, Nemanja discloses:
A route planning system comprising:
a hardware processor (see at least Nemanja, ¶¶ [0050]) configured to implement:
an information acquiring unit (see at least Nemanja, ¶¶ [0009], [0029])
a first processing module (see at least Nemanja, ¶¶ [0010]-[0011])
a second processing module (see at least Nemanja, ¶¶ [0010]-[0011])
the information acquiring unit configured to obtain target information including a start state and a goal state of each of a plurality of agents in a continuous state space (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
wherein the roadmap constructing model includes:
the first processing module configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
the second processing module configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step, and candidate states of the other agents in a time step before the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
estimating one or more candidate states in a next time step using the roadmap constructing model that has completed training by designating the start state of the one agent represented in the obtained target information as a candidate state of the target agent in an initial target time step (see at least Nemanja, ¶¶ [0014]-[0015], [0029]-[0030], [0040]-[0042], [0062]-[0063], which discloses initializing the planning by treating the car’s current state as the starting candidate and using the roadmap module to figure out where it could go in the next time step, this means that one or more candidate states are estimated using the model by designating a start point of the target agent obtained from the target information in an initial target time step)
repeating estimating of the candidate state of a next time step using the roadmap constructing model that has completed training by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent (see at least Nemanja, ¶¶ [0014]-[0015], [0033]-[0034], [0043], [0046]-[0047] which discloses the iterative process of estimating a vehicle’s candidate state through planned movements using a learning model until the vehicle has reached its goal state or near the goal state, this means that the estimate of the candidate state of a next time step using the model by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent is repeated)
a third processing module (see at least Nemanja, ¶¶ [0009], [0029])
the third processing module configured to generate third characteristics information from environmental information including the information relating to obstacles (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
a map constructing unit (see at least Shai, ¶¶ [0144]-[0145])
the map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, and information relating to obstacles present in the continuous state space (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space)
a search unit (see at least Shai, ¶¶ [0519]-[0520])
the search unit configured to search for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent (see at least Shai, ¶¶ [0519]-[0520], [0579] which discloses the system searching for longitudinal and lateral goals of various agents in a state space including the current vehicle, taking into consideration each goal state (route) when mapping, which means a route for each of the agents is searched and utilized to find a more optimal route for the current agent and a start and goal state is constructed for each one)
an estimation module (see at least Shai, ¶¶ [0505]-[0507])
the estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents (see at least Shai, ¶¶ [0009], [0029], [0056]-[0059], [0065] which discloses the system selecting one target agent amongst a plurality of agents, which means executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, a search unit configured to search for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent, an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated, wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning, and wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents as taught by Shai. To note, the examiner does not concede that the reference of Nemanja does not construct maps for navigation, much of the invention’s navigational techniques using machine learning and mapping processes would be redundant to its embodiment. However, the examiner acknowledges that the processes of the roadmap constructing model are more detailed and geared towards the details of the specific claim limitations. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 2, Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
wherein the estimation module is configured to estimate one or more candidate states in the next time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the using of the roadmap constructing model that has completed training includes configuring of the environmental information from the information included in the obtained target information and giving of the configured environmental information to the third processing module (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a third processing module configured to generate third characteristics information from environmental information including information relating to obstacles, wherein the estimation module is configured to estimate one or more candidate states in the next time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated, and wherein the using of the roadmap constructing model that has completed training includes configuring of the environmental information from the information included in the obtained target information and giving of the configured environmental information to the third processing module as taught by Shai. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 5, Nemanja discloses
The route planning system according to claim 1, wherein the attributes of the target agent include at least one of a size, a shape, a maximum speed, and a weight (see at least Nemanja, ¶¶ [0047]-[0048] which discloses various attributes of a target agent such as speed, acceleration, etc.)
Regarding claim 6, Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
The route planning system according to claim 1,wherein the other agent information is configured to further include attributes of the other agents (see at least Shai, ¶¶ [0527]-[0531] which discloses other agent information configured to further include attributes of the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include the route planning system according to claim 1,wherein the other agent information is configured to further include attributes of the other agents as taught by Shai. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 7, Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
The route planning system according to claim 6, wherein the attributes of the other agents include at least one of a size, a shape, a maximum speed, and a weight (see at least Shai, ¶¶ [0527]-[0531] which discloses other agent information configured to further include attributes of the other agents such as speed and type/size)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include The route planning system according to claim 6, wherein the attributes of the other agents include at least one of a size, a shape, a maximum speed, and a weight as taught by Shai. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 8, Nemanja discloses:
The route planning system according to claim 1, wherein the target agent information is configured to further include a direction flag representing a direction in which the target agent is to transition in the continuous state space (see at least Nemanja, ¶¶ [0035]-[0036] which discloses the target agent information including a direction or multiple directions in which the target agent (SDV) is to transition in the continuous (iterative) state space)
Regarding claim 9, Nemanja discloses:
The route planning system according to claim 1,wherein each of the plurality of agents is a mobile body configured to autonomously move (see at least Nemanja, ¶¶ [0067] which discloses where each of the plurality of agents is configured to autonomously move)
Regarding claim 10, Nemanja discloses:
A route planning method causing a computer to execute:
a step of obtaining target information including a start state and a goal state of each of a plurality of agents in a continuous state space (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit performing a step of (sensors deployed on the vehicle) obtaining target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
wherein the roadmap constructing model includes:
a first processing module, implemented by a hardware processor (see at least Nemanja, ¶¶ [0050]) and configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step, a candidate state of the target agent in a time step before the target time step, and attributes of the target agent (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
a second processing module, implemented by a hardware processor and configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step, and candidate states of the other agents in a time step before the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
estimating one or more candidate states in a next time step using the roadmap constructing model that has completed training by designating the start state of the one agent represented in the obtained target information as a candidate state of the target agent in an initial target time step (see at least Nemanja, ¶¶ [0014]-[0015], [0029]-[0030], [0040]-[0042], [0062]-[0063], which discloses initializing the planning by treating the car’s current state as the starting candidate and using the roadmap module to figure out where it could go in the next time step, this means that one or more candidate states are estimated using the model by designating a start point of the target agent obtained from the target information in an initial target time step)
repeating estimating of the candidate state of a next time step using the roadmap constructing model that has completed training by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent (see at least Nemanja, ¶¶ [0014]-[0015], [0033]-[0034], [0043], [0046]-[0047] which discloses the iterative process of estimating a vehicle’s candidate state through planned movements using a learning model until the vehicle has reached its goal state or near the goal state, this means that the estimate of the candidate state of a next time step using the model by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent is repeated)
a third processing module (see at least Nemanja, ¶¶ [0009], [0029])
the third processing module, implemented by the hardware processor and configured to generate third characteristics information from environmental information including the information relating to obstacles (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
a step of constructing a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, and information relating to obstacles present in the continuous state space (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space)
a step of searching for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent (see at least Shai, ¶¶ [0519]-[0520], [0579] which discloses the system searching for longitudinal and lateral goals of various agents in a state space including the current vehicle, taking into consideration each goal state (route) when mapping, which means a route for each of the agents is searched and utilized to find a more optimal route for the current agent and a start and goal state is constructed for each one)
an estimation module implemented by the hardware processor and configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents (see at least Shai, ¶¶ [0009], [0029], [0056]-[0059], [0065] which discloses the system selecting one target agent amongst a plurality of agents, which means executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a step of constructing a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, a step of searching for a route of each of the agents from the start state to the goal state on the roadmap constructed for each agent, an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated, and wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning as taught by Shai. To note, the examiner does not concede that the reference of Nemanja does not construct maps for navigation, much of the invention’s navigational techniques using machine learning and mapping processes would be redundant to its embodiment. However, the examiner acknowledges that the processes of the roadmap constructing model are more detailed and geared towards the details of the specific claim limitations. Additionally, the embodiment of Nemanja is capable to of estimating/predicting the candidate states of the target vehicle. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 11, Nemanja discloses:
A roadmap constructing device comprising:
a hardware processor (see at least Nemanja, ¶¶ [0050]) configured to implement:
an information acquiring unit (see at least Nemanja, ¶¶ [0009], [0029])
a first processing module (see at least Nemanja, ¶¶ [0010]-[0011])
a second processing module (see at least Nemanja, ¶¶ [0010]-[0011])
the information acquiring unit configured to obtain target information including a start state and a goal state of each of a plurality of agents in a continuous state space (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
wherein the roadmap constructing model includes:
the first processing module configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step, a candidate state of the target agent in a time step before the target time step, and attributes of the target agent (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
the second processing module configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step, and candidate states of the other agents in a time step before the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
estimating one or more candidate states in a next time step using the roadmap constructing model that has completed training by designating the start state of the one agent represented in the obtained target information as a candidate state of the target agent in an initial target time step (see at least Nemanja, ¶¶ [0014]-[0015], [0029]-[0030], [0040]-[0042], [0062]-[0063], which discloses initializing the planning by treating the car’s current state as the starting candidate and using the roadmap module to figure out where it could go in the next time step, this means that one or more candidate states are estimated using the model by designating a start point of the target agent obtained from the target information in an initial target time step)
repeating estimating of the candidate state of a next time step using the roadmap constructing model that has completed training by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent (see at least Nemanja, ¶¶ [0014]-[0015], [0033]-[0034], [0043], [0046]-[0047] which discloses the iterative process of estimating a vehicle’s candidate state through planned movements using a learning model until the vehicle has reached its goal state or near the goal state, this means that the estimate of the candidate state of a next time step using the model by designating each of the one or more estimated candidate states in the next time step as a candidate state of a new target time step until the goal state or a state near the goal state of the one agent is included in estimated one or more candidate states in the next time step by designating each of the plurality of agents as the target agent is repeated)
a third processing module (see at least Nemanja, ¶¶ [0009], [0029])
the third processing module configured to generate third characteristics information from environmental information including the information relating to obstacles (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
a map constructing unit (see at least Nemanja, ¶¶ [0144]-[0145])
the map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, and information relating to obstacles present in the continuous state space (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space)
an estimation module (see at least Shai, ¶¶ [0505]-[0507])
the estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents (see at least Shai, ¶¶ [0009], [0029], [0056]-[0059], [0065] which discloses the system selecting one target agent amongst a plurality of agents, which means executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a map constructing unit configured to construct a roadmap for each of the agents from the obtained target information using a roadmap constructing model that has completed training, an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated, wherein the roadmap constructing model that has completed training is generated using machine learning using learning data acquired from correct answer routes of a plurality of agents for learning, and wherein the constructing of the roadmap for each agent is configured by executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents as taught by Shai. To note, the examiner does not concede that the reference of Nemanja does not construct maps for navigation, much of the invention’s navigational techniques using machine learning and mapping processes would be redundant to its embodiment. However, the examiner acknowledges that the processes of the roadmap constructing model are more detailed and geared towards the details of the specific claim limitations. Additionally, the embodiment of Nemanja is capable to of estimating/predicting the candidate states of the target vehicle. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 12, Nemanja discloses:
A model generating device comprising:
a hardware processor (see at least Nemanja, ¶¶ [0050]) configured to implement:
a data acquiring unit (see at least Nemanja, ¶¶ [0009], [0029])
a learning processing unit (see at least Nemanja, ¶¶ [0010]-[0011])
a first processing module (see at least Nemanja, ¶¶ [0010]-[0011])
a second processing module (see at least Nemanja, ¶¶ [0010]-[0011])
the learning processing unit configured to perform machine learning of a roadmap constructing model using the obtained learning data (see at least Nemanja, ¶¶ [0010]-[0011] which discloses a machine learning model to perform machine learning, this means that the system implements machine learning of a roadmap constructing model using the obtained machine learning data)
wherein the roadmap constructing model includes:
the first processing module configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step in a continuous state space, a candidate state of the target agent in a time step before the target time step, and attributes of the target agent (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
the second processing module configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step, and candidate states of the other agents in a time step before the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
training the roadmap constructing model such that a candidate state of the target agent in a next time step, which is estimated by the estimation module, is appropriate for a state of the one agent for learning in the second time step by, for each data set (see at least Nemanja, ¶¶ [0014]-[0015], [0029]-[0030], [0040]-[0042], [0062]-[0063], which discloses initializing the planning by treating the car’s current state as the starting candidate and using the roadmap module to figure out where it could go in the next time step, this means that one or more candidate states are estimated using the model by designating a start point of the target agent obtained from the target information in an initial target time step)
giving a state of the one agent for learning in the first time step to the first processing module as a candidate state of the target agent in the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
giving states of at least some of remaining agents for learning in the first time step to the second processing module as candidate states of the other agents in the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
a third processing module (see at least Nemanja, ¶¶ [0009], [0029])
the third processing module configured to generate third characteristics information from environmental information including the information relating to obstacles (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
a data acquiring unit configured to obtain learning data generated from correct answer routes of a plurality of agents for learning (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
an estimation module (see at least Shai, ¶¶ [0505]-[0507])
the estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information, the second characteristics information, and the third characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the learning data includes a goal state in the correct answer route of each agent for learning and a plurality of data sets, wherein each of the plurality of data sets is configured using a combination of a state of each agent for learning in a first time step and a state of each agent for learning in a second time step (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, including a first time step (current, or future) and second time step (first time step becomes current, second time step occurs successively after), [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning)
wherein the second time step is a time step next to the first time step (see at least Shai, ¶¶ [0144]-[0145] which discloses where the second time step is a time step next to the first time step), [0180] discloses an example where a current location/movement of a first vehicle occurs followed by a second time step of a future path ahead of the first location)
wherein the machine learning of the roadmap constructing model is configured by:
handling one agent for learning among the plurality of agents for learning as the target agent; handling at least some of remaining agents for learning among the plurality of agents for learning as the other agents (see at least Shai, ¶¶ [0009], [0029], [0056]-[0059], [0065] which discloses the system selecting one target agent amongst a plurality of agents, which means executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a data acquiring unit configured to obtain learning data generated from correct answer routes of a plurality of agents for learning, an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated, wherein the learning data includes a goal state in the correct answer route of each agent for learning and a plurality of data sets, wherein each of the plurality of data sets is configured using a combination of a state of each agent for learning in a first time step and a state of each agent for learning in a second time step, wherein the second time step is a time step next to the first time step, and handling one agent for learning among the plurality of agents for learning as the target agent; handling at least some of remaining agents for learning among the plurality of agents for learning as the other agents as taught by Shai. To note, the examiner does not concede that the reference of Nemanja does not construct maps for navigation, much of the invention’s navigational techniques using machine learning and mapping processes would be redundant to its embodiment. However, the examiner acknowledges that the processes of the roadmap constructing model are more detailed and geared towards the details of the specific claim limitations. Additionally, the embodiment of Nemanja is capable to of estimating/predicting the candidate states of the target vehicle. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 13, Nemanja discloses:
The model generating device according to wherein the target agent information is configured to:
further include a direction flag representing a direction in which the target agent is to transition in the continuous state space (see at least Nemanja, ¶¶ [0035]-[0036] which discloses the target agent information including a direction or multiple directions in which the target agent (SDV) is to transition in the continuous (iterative) state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
wherein each data set is configured to further include a training flag representing a direction from the state of the first time step to the state of the second time step in the continuous state space (see at least Shai, ¶¶ [0199]-[0203], [0326], [0519]-[0520], which discloses generating a characteristic vector that indicates whether the target agent (current vehicle) becomes closer to the goal state in response to a particular observed state, this means that each data set is configured to include a form of training flag representing a direction from the state of the first time step to the state of the second time step in a continuous state space)
wherein the machine learning of the roadmap constructing model includes giving of the training flag of the one agent for learning to the first processing module as a direction flag of the target agent when a candidate state of the target agent in a next time step is estimated for each data set (see at least Shai, ¶¶ [0199]-[0203], [0326], [0519]-[0520], which discloses generating a characteristic vector that indicates whether the target agent (current vehicle) becomes closer to the goal state in response to a particular observed state, this means that each data set is configured to include a form of training flag representing a direction from the state of the first time step to the state of the second time step in a continuous state space using [0525] machine learning)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include wherein each data set is configured to further include a training flag representing a direction from the state of the first time step to the state of the second time step in the continuous state space and wherein the machine learning of the roadmap constructing model includes giving of the training flag of the one agent for learning to the first processing module as a direction flag of the target agent when a candidate state of the target agent in a next time step is estimated for each data set as taught by Shai. Incorporating the teachings of Shai into the base device of Nemanja would allow for an improvement geared towards a smoother iterative process and further adaptability.
Regarding claim 14, Nemanja discloses:
A model generating method causing a computer to execute:
a step of performing machine learning of a roadmap constructing model using the obtained learning data (see at least Nemanja, ¶¶ [0010]-[0011] which discloses a step of performing machine learning)
wherein the roadmap constructing model includes:
a first processing module implemented by a hardware processor and configured to generate first characteristics information from target agent information including a goal state of a target agent and a candidate state of a target time step in a continuous state space, a candidate state of the target agent in a time step before the target time step, and attributes of the target agent (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step, [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space))
a second processing module implemented by the hardware processor and configured to generate second characteristics information from other agent information including goal states of other agents other than the target agent and candidate states of the target time step, and candidate states of the other agents in a time step before the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
training the roadmap constructing model such that a candidate state of the target agent in a next time step, which is estimated by the estimation module, is appropriate for a state of the one agent for learning in the second time step by, for each data set (see at least Nemanja, ¶¶ [0014]-[0015], [0029]-[0030], [0040]-[0042], [0062]-[0063], which discloses initializing the planning by treating the car’s current state as the starting candidate and using the roadmap module to figure out where it could go in the next time step, this means that one or more candidate states are estimated using the model by designating a start point of the target agent obtained from the target information in an initial target time step)
giving a state of the one agent for learning in the first time step to the first processing module as a candidate state of the target agent in the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0036] which discloses generating first characteristics information such as a current vehicle (SDV) state to an endpoint in a given region, and this means that the first processing device generates first characteristics information including target agent information (SDV) such as a goal and candidate state of a target time step)
giving states of at least some of remaining agents for learning in the first time step to the second processing module as candidate states of the other agents in the target time step (see at least Nemanja, ¶¶ [0010]-[0011], [0037], [0056]-[0059], [0065] which discloses second characteristics information from other vehicles and potential hazards, or objects of interest like current heading, this means that a second processing module generates second characteristics information from other agent information (other vehicles, obstacles, etc.) of the target time step)
a third processing module configured to generate third characteristics information from environmental information including the information relating to obstacles (see at least Nemanja, ¶¶ [0009], [0029], [0056]-[0059], [0065]which discloses an information acquiring unit (sensors deployed on the vehicle) configured to obtain target information such as a current state (current location and heading of a vehicle) and the goal state (destination, planned route endpoint) of each of a plurality of agents (current vehicle, obstacles, pedestrians, other vehicles, etc.) in a continuous state space)
Nemanja is silent on, however, in the same field of endeavor, Shai teaches:
a step of obtaining learning data generated from correct answer routes of a plurality of agents for learning (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
an estimation module implemented by the hardware processor and configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information, the second characteristics information, and the tird characteristics information that are generated (see at least Shai, ¶¶ [0505]-[0507] which discloses estimating one or more current and future states of a current agent (current vehicle) relative to its current state and that of other agents (vehicles) such as sensed actions and positioning information, this means that the system is able to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated)
wherein the learning data includes a goal state in the correct answer route of each agent for learning and a plurality of data sets (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning )
wherein each of the plurality of data sets is configured using a combination of a state of each agent for learning in a first time step and a state of each agent for learning in a second time step (see at least Shai, ¶¶ [0144]-[0145] which discloses a system that is configured to construct a 3D roadmap for each of the agents (current vehicle and its surroundings, other vehicles, pedestrians, obstacles, etc.) using a learning model, which means that a deep learning model that has completed training constructs the map based on the obtained target information from each of the agents in the state space, including a first time step (current, or future) and second time step (first time step becomes current, second time step occurs successively after), [0180] discloses that the analysis may be performed by a learning model using machine learning to achieve optimal solutions, this means that the model utilizes learning data acquired from correct answer routes of a plurality of agents for learning)
wherein the second time step is a time step next to the first time step (see at least Shai, ¶¶ [0144]-[0145] which discloses where the second time step is a time step next to the first time step), [0180] discloses an example where a current location/movement of a first vehicle occurs followed by a second time step of a future path ahead of the first location)
wherein the machine learning of the roadmap constructing model is configured by:
handling one agent for learning among the plurality of agents for learning as the target agent and handling at least some of remaining agents for learning among the plurality of agents for learning as the other agents (see at least Shai, ¶¶ [0009], [0029], [0056]-[0059], [0065] which discloses the system selecting one target agent amongst a plurality of agents, which means executing a process of handling one agent among the plurality of agents as the target agent, handling at least some of remaining agents among the plurality of agents as the other agents)
It would have been obvious to a person of ordinary skill in the art to modify Nemanja to include a step of obtaining learning data generated from correct answer routes of a plurality of agents for learning, an estimation module configured to estimate one or more candidate states of the target agent in a time step next to the target time step from the first characteristics information and the second characteristics information that are generated, wherein the learning data includes a goal state in the correct answer route of each agent for learning and a plurality of data sets, wherein each of the plurality of data sets is configured using a combination of a state of each agent for learning in a first time step and a state of each agent for learning in a second time step, wherein the second time step is a time step next to the first time step, and wherein the machine learning of the roadmap constructing model is configured by: handling one agent for learning among the plurality of agents for learning as the target agent and handling at least some of remaining agents for learning among the plurality of agents for learning as the other agents as taught by Shai. Incorporating the teachings of Shai into Nemanja would allow for an improvement to the base invention of Nemanja that further advances the reliability and adaptability of its navigation method to more dynamic scenarios regarding a multitude of dynamic agents in a vehicle controls environment.
Regarding claim 19, Nemanja discloses:
The route planning system according to claim 2,wherein the attributes of the target agent include at least one of a size, a shape, a maximum speed, and a weight (see at least Nemanja, ¶¶ [0047]-[0048] which discloses various attributes of a target agent such as speed, acceleration, etc.)
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 KIRSTEN JADE M SANTOS whose telephone number is (571)272-7442. The examiner can normally be reached Monday - Friday: 9:00 am - 5:00 pm, (+ with flex).
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, Rachid Bendidi can be reached at (571) 272-4896. 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.
/KIRSTEN JADE M SANTOS/Examiner, Art Unit 3664
/RACHID BENDIDI/Supervisory Patent Examiner, Art Unit 3664