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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 04/01/2026 has been entered.
Status of the Claims
This action is in response to the applicant’s amendment/response of March 20, 2026 and RCE of April 1, 2026.
Claims 34 and 35 have been newly added.
Claims 1-21, 34, and 35 are pending and have been considered as follows.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on January 13, 2025, March 26, 2026, and April 13, 2026. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Arguments
Applicant’s arguments/amendments with respect to the objection to the claims have been fully considered and are persuasive. Therefore, the objection to the claims has been withdrawn.
Applicant’s arguments/amendments with respect to the rejection of claims under 35 USC § 103 have been fully considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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.
Claim(s) 1-5, 7-12, 15-17, 20, 21, 34, and 35 are rejected under 35 U.S.C. 103 as being unpatentable over Imam et al., US 2018/0174041 A1, hereinafter referred to as Imam, in view of MORGAN-BROWN, US 2019/0316924 A1, hereinafter referred to as MORGAN-BROWN, and further in view of Park et al., US 2006/0149465 A1, hereinafter referred to as Park, respectively.
As to claim 1, Imam teaches one or more processors, comprising:
circuitry to (see at least paragraphs 27 and 36-38, Imam):
control one or more functionalities of an autonomous device to cause the autonomous device to traverse one or more paths among a plurality of paths in an environment (see at least paragraphs 54-63, 68-70, and 107 regarding neurons (e.g., 651-656) of the SNN used to model a path between points in a network (through a spike chain generated through activation of a source neuron (e.g., 652) on synapses (e.g., 657-661) strengthened through previous training of the SNN)) may include further synapses to send output spikes (e.g., 665) to motor control logic 650 to direct an autonomous passenger vehicle, drone, robot, etc. to navigate the physical space corresponding to the SNN. Using a configurable SNN to determine paths between nodes in a particular network topology may be leveraged to implement SNNs to realize hardware-based path planning in physical space. For instance, as shown in the simplified block diagram 700 of FIG. 7, path planning may involve making a navigation decision at a particular starting point (e.g., 705) in a physical environment, where the goal of the navigation is to find an optimized path to a desired destination, or reward location (e.g., 710), Imam), wherein the circuitry is further to:
use the one or more neural networks to calculate the plurality of paths from the different starting locations to the one or more destination locations, through which the autonomous device is to traverse, based, at least in part, on the one or more outputs (see at least FIG. 2A and paragraphs 21-27. See also at least FIGS. 5A-5F and paragraphs 54-63 regarding determining a shortest path between any two nodes in a network modeled by the SNN. A controller utility (internal or external to the neuromorphic computing device) may be associated with the monitoring of traffic in the SNN, such that the monitoring is orchestrated in connection with the training of the SNN and activation of the designated source neuron (as well as the activation of the designated destination neuron during training). Further, following training the SNN to determine shortest paths to a particular node (and corresponding neuron) in a network, additional, alternative source nodes may also be tested to determine shortest paths from these other source nodes (effectively reusing the training of the SNN corresponding to the designation of the particular node (e.g., Neuron 6) as the destination node). It may be desirable to use an SNN implemented to model a particular network topology to determine shortest paths from one or more source nodes to another destination node. See also at least paragraphs 68-69 regarding neurons (e.g., 651-656) of the SNN used to model a path between points in a network (through a spike chain generated through activation of a source neuron (e.g., 652) on synapses (e.g., 657-661) strengthened through previous training of the SNN)) may include further synapses to send output spikes (e.g., 665) to motor control logic 650 to direct an autonomous passenger vehicle, drone, robot, etc. to navigate the physical space corresponding to the SNN. Path planning may involve making a navigation decision at a particular starting point (e.g., 705) in a physical environment, where the goal of the navigation is to find an optimized path to a desired destination, or reward location (e.g., 710). In such examples, a SNN may be implemented using a neuromorphic computing device, with each neuron in the SNN assigned to correspond to a point within the physical environment. As in the examples of FIGS. 4A-5H, a destination neuron (corresponding to a physical destination) may be selected and activated to train the SNN to bias subsequent spike chains to be directed along a shortest path toward the destination, Imam).
Imam does not explicitly teach causing one or more neural networks to model the environment comprising one or more encoded values indicating one or more distances for each starting location of different starting locations to one or more destination locations in the environment; or generating one or more outputs including the one or more distances from the different starting locations to the one or more destination locations.
However, MORGAN-BROWN teaches causing one or more neural networks to model the environment comprising one or more encoded values indicating one or more distances for each starting location of different starting locations to one or more destination locations in the environment (see at least FIGS. 2D-2E and paragraphs 47-48 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in FIG. 2E. See also at least paragraphs 94-114 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations. A neural network is trained on real cost parameter data for example as follows: Features (the input for training the model): source latitude in radians, source longitude in radians, target latitude in radians, target longitude in radians, haversine distance between source and target (which may be scaled to be between 0 and 1 by dividing by the maximum theoretical range of the vehicle). Labels (the expected output of the trained model): the ratio of the real distance to the haversine distance, the ratio of the real time to the haversine distance, the ratio of the real elevation change to the haversine distance, the ratio of other cost parameters to the haversine distance. With a suitable loss function (such as mean squared error) an optimizer (such as Stochastic Gradient Descent (SGD) or Adam) can then be used to fit the neural network. Once the neural network has been trained in this way it can be deployed into the vehicle and used to estimate the cost parameters. See also at least Claim 38); and generating one or more outputs including the one or more distances from the different starting locations to the one or more destination locations (see at least FIGS. 2D-2E and paragraphs 47-48. See also at least paragraphs 94-114 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations. A neural network is trained on real cost parameter data for example as follows: Features (the input for training the model): source latitude in radians, source longitude in radians, target latitude in radians, target longitude in radians, haversine distance between source and target (which may be scaled to be between 0 and 1 by dividing by the maximum theoretical range of the vehicle). Labels (the expected output of the trained model): the ratio of the real distance to the haversine distance, the ratio of the real time to the haversine distance, the ratio of the real elevation change to the haversine distance, the ratio of other cost parameters to the haversine distance. With a suitable loss function (such as mean squared error) an optimizer (such as Stochastic Gradient Descent (SGD) or Adam) can then be used to fit the neural network. Once the neural network has been trained in this way it can be deployed into the vehicle and used to estimate the cost parameters. See also at least Claim 38 regarding cost parameters between locations for a particular geographical area and using these to calculate a set of weights that can be applied to computed haversine distances to create a prediction of cost parameters).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches causing one or more neural networks to model the environment comprising one or more encoded values indicating one or more distances for each starting location of different starting locations to one or more destination locations in the environment; and generating one or more outputs including the one or more distances from the different starting locations to the one or more destination locations with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of causing one or more neural networks to model the environment comprising one or more encoded values indicating one or more distances for each starting location of different starting locations to one or more destination locations in the environment; and generating one or more outputs including the one or more distances from the different starting locations to the one or more destination locations and would have predictably applied it to improve the system of Imam.
Imam, as modified by MORGAN-BROWN, does not explicitly teach wherein the different starting locations along the one or more paths are to be selected such that the one or more distances between the different starting locations and the one or more destination locations successively decrease in length as the autonomous device traverses the one or more paths.
However, such matter is taught by Park (see at least Abstract regarding planning the minimum cost path from the movement path by selecting one or more shortest-distance via points from the via points. See also at least FIG. 2A-2C and paragraphs 38-41 regarding a start point S has a horizontal coordinate of 1 and a vertical coordinate of 3, and is thus expressed as Cell(1,3). In a case in which a cost needed to move from each cell to the goal is expressed as "move_cost", the move cost at Cell(1,3) is expressed as move_cost(Cell(1,3))", and the move cost has a value of 15. The path may be set by selecting a subsequent cell following a current cell. Cell(1,2), Cell(1,4), and Cell(2,2) each neighbor the cell with the start point S, i.e., Cell(1,3). Among those cells, Cell(2,2) has the lowest move cost. After moving to Cell(2,2), a cell having a lowest move cost is selected from among the cells neighboring Cell(2,2). Cell(1,2), Cell(1,1), Cell(2,1), Cell(3,1), and Cell(3,2) each neighbor Cell(2,2), and have weights 14(Cell(1,2)), 15(Cell(1,1)), 14(Cell(2,1)), 13(Cell(3,1)), and 12(Cell(3,2)). Among those cells, Cell(3,2), which has the lowest move cost, is selected).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Park which teaches wherein the different starting locations along the one or more paths are to be selected such that the one or more distances between the different starting locations and the one or more destination locations successively decrease in length as the autonomous device traverses the one or more paths with the system of Imam, as modified by MORGAN-BROWN, as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of wherein the different starting locations along the one or more paths are to be selected such that the one or more distances between the different starting locations and the one or more destination locations successively decrease in length as the autonomous device traverses the one or more paths and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN.
As to claim 2, Imam teaches obtaining a first location and receiving, as input to the one or more neural networks, the different starting locations and the one or more destination locations (see at least FIGS. 5A-5F and paragraphs 54-63 regarding the block diagram 500b of FIG. 5B illustrates a simplified SNN including ten bi-directionally connected neurons, each neuron representing a respective node in a particular network. Neighboring nodes or connections between nodes in the particular network may be modeled through connections, or synapses, between the artificial neurons modeling the nodes in the SNN. In the example of FIG. 5F, Neuron 1 is selected to designate a corresponding node as the source node and define the problem statement: determining a path in a particular network from a node represented by Neuron 1 and a destination node represented (in this example) by Neuron 6. In other transactions, a different one of the neurons (i.e., other than Neuron 6) may be designated as the destination and other neurons may designated as the source, and so on. To designate the source neuron (e.g., Neuron 1) following training of the SNN (e.g., through the initial spike wave and corresponding STDP adjustment and increase in spiking threshold within the SNN (as discussed in FIGS. 5C-5E)), the designated source neuron may be activated (e.g., at 515) in a manner similar to the activation (e.g., 510) of the destination neuron (e.g., in FIG. 5C), Imam); and
calculating the plurality of paths (see at least FIGS. 5A-5F and paragraphs 54-63 regarding the spike chain from source Neuron 1 to Neuron 2 to Neuron 4 to the destination Neuron 6, may be interpreted to represent the shortest path between the source Neuron 1 and the destination Neuron 6. Monitoring and/or reporting logic may be provided to identify the order ad source of spikes sent in this chain of spikes. For instance, routers of a neuromorphic computing device may be monitored to identify traffic between (or within) one or more neuromorphic cores that identifies the spike messages that result from activating Neuron 1 following training of the SNN to determine shortest paths across a network to a designated destination node, Imam).
Imam does not explicitly teach causing the one or more neural networks to output the one or more distances from the different starting locations and the one or more destination locations.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-35 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. See also at least paragraphs 80-86 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches causing the one or more neural networks to output the one or more distances from the different starting locations and the one or more destination locations with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of causing the one or more neural networks to output the one or more distances from the different starting locations and the one or more destination locations and would have predictably applied it to improve the system of Imam.
As to claim 3, Imam teaches wherein: the first location is a location of the autonomous device (see at least paragraphs 60-69 regarding path planning may involve making a navigation decision at a particular starting point (e.g., 705) in a physical environment, where the goal of the navigation is to find an optimized path to a desired destination, or reward location (e.g., 710). In such examples, a SNN may be implemented using a neuromorphic computing device, with each neuron in the SNN assigned to correspond to a point within the physical environment. As in the examples of FIGS. 4A-5H, a destination neuron (corresponding to a physical destination) may be selected and activated to train the SNN to bias subsequent spike chains to be directed along a shortest path toward the destination. For instance, a destination neuron may be activated that corresponds to a reward location (e.g., a destination of a delivery drone's flight, a charging station for a robotic device, a location of an object to be retrieved by an autonomous tool, etc.), resulting in a spike wave (e.g., 715) that propagates from the destination neuron 710 to all other neurons in the SNN. As in the examples of FIGS. 5B-5H, an STDP rule may be applied in the SNN to cause synapses directed toward the destination neuron to be strengthened based on the spike wave. Further, a spiking threshold potential may be adjusted to further bias the trained SNN. A current location of a device (or individual wearing or carrying the device) may be determined and this location may be the basis of activating another one of the neurons in the SNN as the source neuron in the path, Imam); and
a subset of locations of the different starting locations are accessible to the autonomous device from the first location (see at least FIGS. 5A-5F and paragraphs 54-63 regarding Neuron 1 may be activated following training of the SNN causing a spike to be sent from Neuron 1 to Neuron 2 at time t=0. A spike may also be sent by Neuron 1 to other neighboring neurons (e.g., Neuron 3), but due to the increased threshold potential applied across the SNN, if the synapse connecting Neuron 1 to these other neighboring neurons (e.g., Neuron 3) was not strengthened according to the STDP during training, the spike will be insufficient to meet the newly increased spiking threshold potential of the receiving neuron and that spike will not propagate further, Imam).
As to claim 4, Imam does not explicitly teach obtaining a subset of distances of the one or more distances corresponding to the subset of locations; selecting a second location of the subset of locations based at least in part on the subset of distances; or calculating the first path comprising a path from the first location to the second location.
However, MORGAN-BROWN teaches obtaining a subset of distances of the one or more distances corresponding to the subset of locations (see at least FIGS. 2D-2E and paragraphs 34-39); selecting a second location of the subset of locations based at least in part on the subset of distances (see at least FIGS. 2D-2E and paragraphs 34-39 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. The calculating comprises identifying which path traversals are on the lowest cost route from the start location node to each other node in the node network); and calculating the first path comprising a path from the first location to the second location (see at least FIGS. 2D-2E and paragraphs 34-39 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. The calculating comprises identifying which path traversals are on the lowest cost route from the start location node to each other node in the node network).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches obtaining a subset of distances of the one or more distances corresponding to the subset of locations; selecting a second location of the subset of locations based at least in part on the subset of distances; and calculating the first path comprising a path from the first location to the second location with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of obtaining a subset of distances of the one or more distances corresponding to the subset of locations; selecting a second location of the subset of locations based at least in part on the subset of distances; and calculating the first path comprising a path from the first location to the second location and would have predictably applied it to improve the system of Imam.
As to claim 5, Imam does not explicitly teach wherein the second location corresponds to a minimum distance of the subset of distances.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-39 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. The calculating comprises identifying which path traversals are on the lowest cost route from the start location node to each other node in the node network).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches wherein the second location corresponds to a minimum distance of the subset of distances with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of having wherein the second location corresponds to a minimum distance of the subset of distances and would have predictably applied it to improve the system of Imam.
As to claim 7, Imam does not explicitly teach wherein a distance of the one or more distances corresponds to a path from a location of the different starting locations to the one or more destination locations.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-35 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. See also at least paragraphs 80-86 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches wherein a distance of the one or more distances corresponds to a path from a location of the different starting locations to the one or more destination locations with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of having wherein a distance of the one or more distances corresponds to a path from a location of the different starting locations to the one or more destination locations and would have predictably applied it to improve the system of Imam.
As to claim 8, Examiner notes claim 8 recites similar limitations to claim 1 and is rejected under the same rational.
As to claim 9, Imam teaches obtaining features of the environment (see at least paragraphs 21-23 regarding such devices 110a-c may detect and/or measure attributes of an environment and generate sensor data describing or capturing characteristics of the environment. See also at least paragraphs 68-70); and select a location in the environment (see at least paragraphs 68-70 regarding a current location of a device (or individual wearing or carrying the device) may be determined and this location may be the basis of activating another one of the neurons in the SNN as the source neuron in the path. As described above, activating the source neuron in the trained SNN may cause a spike chain to propagate along a shortest path through the SNN to the destination. A corresponding physical location may be determined for each neuron sending a spike in the spike chain to determine a shortest path to the physical destination environment (e.g., by navigating, in order, through each one of the locations corresponding to the neurons (e.g., Neurons A-F) in the spike chain)).
Imam does not explicitly teach inputting at least the features and the location to the one or more neural networks to obtain the one or more distances.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-35 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. See also at least paragraphs 80-86 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches inputting at least the features and the location to the one or more neural networks to obtain the one or more distances with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of inputting at least the features and the location to the one or more neural networks to obtain the one or more distances and would have predictably applied it to improve the system of Imam.
As to claim 10, Imam teaches selecting the different starting locations accessible to the autonomous device (see at least FIGS. 5A-5F and paragraphs 54-63 regarding Neuron 1 may be activated following training of the SNN causing a spike to be sent from Neuron 1 to Neuron 2 at time t=0. A spike may also be sent by Neuron 1 to other neighboring neurons (e.g., Neuron 3), but due to the increased threshold potential applied across the SNN, if the synapse connecting Neuron 1 to these other neighboring neurons (e.g., Neuron 3) was not strengthened according to the STDP during training, the spike will be insufficient to meet the newly increased spiking threshold potential of the receiving neuron and that spike will not propagate further).
Imam does not explicitly teach selecting a first location of the different starting locations based at least in part on the one or more distances, wherein a first path of the plurality of paths indicates a path from the autonomous device to the first location.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-39 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. The calculating comprises identifying which path traversals are on the lowest cost route from the start location node to each other node in the node network).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches selecting a first location of the different starting locations based at least in part on the one or more distances, wherein a first path of the plurality of paths indicates a path from the autonomous device to the first location with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of selecting a first location of the different starting locations based at least in part on the one or more distances, wherein a first path of the plurality of paths indicates a path from the autonomous device to the first location and would have predictably applied it to improve the system of Imam.
As to claim 11, Imam teaches wherein the set of instructions further comprise instructions, which if performed by the one or more processors, cause the one or more processors to cause the autonomous device to navigate to the first location using the first path (see at least paragraphs 54-63, 68-70, and 107 regarding neurons (e.g., 651-656) of the SNN used to model a path between points in a network (through a spike chain generated through activation of a source neuron (e.g., 652) on synapses (e.g., 657-661) strengthened through previous training of the SNN)) may include further synapses to send output spikes (e.g., 665) to motor control logic 650 to direct an autonomous passenger vehicle, drone, robot, etc. to navigate the physical space corresponding to the SNN. Using a configurable SNN to determine paths between nodes in a particular network topology may be leveraged to implement SNNs to realize hardware-based path planning in physical space. For instance, as shown in the simplified block diagram 700 of FIG. 7, path planning may involve making a navigation decision at a particular starting point (e.g., 705) in a physical environment, where the goal of the navigation is to find an optimized path to a desired destination, or reward location (e.g., 710). See also at least paragraph 157 regarding sending a third signal to a controller of an autonomous device to cause the autonomous to navigate a physical path from the starting location to the destination location based on the path, Imam).
As to claim 12, Imam teaches wherein the autonomous device is an autonomous car (see at least paragraph 21 regarding machine learning functionality of the neuromorphic computing system 105 may be consumed by robots, autonomous vehicles, autonomous devices, among other examples, Imam).
As to claim 15, Examiner notes claim 15 recites similar limitations to claim 1 and is rejected under the same rational.
As to claim 16, Imam teaches wherein the one or more processors are further to:
capture a representation of the environment (see at least paragraph 21-23 regarding such devices 110a-c may detect and/or measure attributes of an environment and generate sensor data describing or capturing characteristics of the environment, Imam); and
use the one or more neural networks to calculate the plurality of paths from a first location of the autonomous device to a second location in the environment (see at least FIGS. 5A-5F and paragraphs 54-63 regarding determining a shortest path between any two nodes in a network modeled by the SNN. Using an SNN implemented to model a particular network topology to determine shortest paths from one or more source nodes to another destination node, Imam).
As to claim 17, Imam does not explicitly teach wherein the one or more processors are further to use the one or more neural networks to calculate the one or more distances in the environment based at least in part on the representation of the environment.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-35 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. See also at least paragraphs 80-86 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches using the one or more neural networks to calculate the one or more distances in the environment based at least in part on the representation of the environment with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of using the one or more neural networks to calculate the one or more distances in the environment based at least in part on the representation of the environment and would have predictably applied it to improve the system of Imam.
As to claim 20, Imam teaches wherein the representation of the environment is a 2D or 3D representation (see at least paragraph 107 regarding the network modeled by the SNN may be a network that itself models a physical 2D or 3D environment, Imam).
As to claim 21, Imam teaches wherein the autonomous device is an autonomous robot (see at least paragraph 21 regarding machine learning functionality of the neuromorphic computing system 105 may be consumed by robots, autonomous vehicles, autonomous devices, among other examples, Imam).
As to claim 34, Imam, as modified by MORGAN-BROWN, does not explicitly teach causing the autonomous device to move from a first starting location to a second starting location; selecting a first path from the plurality of paths comprising a shortest distance between the second starting location and a first destination location; causing the autonomous device to move from the second starting location to a third starting location; or selecting a second path from the plurality of paths comprising a shortest distance between the third starting location and the first destination location.
However, Park teaches causing the autonomous device to move from a first starting location to a second starting location (see at least FIG. 2A-2C and paragraphs 38-41 regarding the path may be set by selecting a subsequent cell following a current cell. Cell(1,2), Cell(1,4), and Cell(2,2) each neighbor the cell with the start point S, i.e., Cell(1,3). Among those cells, Cell(2,2) has the lowest move cost. After moving to Cell(2,2)); selecting a first path from the plurality of paths comprising a shortest distance between the second starting location and a first destination location (see at least FIG. 2A-2C and paragraphs 38-41 regarding after moving to Cell(2,2), a cell having a lowest move cost is selected from among the cells neighboring Cell(2,2). Cell(1,2), Cell(1,1), Cell(2,1), Cell(3,1), and Cell(3,2) each neighbor Cell(2,2), and have weights 14(Cell(1,2)), 15(Cell(1,1)), 14(Cell(2,1)), 13(Cell(3,1)), and 12(Cell(3,2)). Among those cells, Cell(3,2), which has the lowest move cost, is selected); causing the autonomous device to move from the second starting location to a third starting location (see at least FIG. 2A-2C and paragraphs 38-41. See also at least FIGS. 7A-8B and paragraphs 70-74 regarding while a mobile home appliance departing from Cell(1,3) is moving toward Cell(8,5) via Cell(2,2) and Cell(3,2), it senses the obstacle upon reaching a position of Cell(5,3)); and selecting a second path from the plurality of paths comprising a shortest distance between the third starting location and the first destination location (see at least FIG. 2A-2C and paragraphs 38-41. See also at least FIGS. 7A-8B and paragraphs 70-74 regarding when replanning the minimum cost path, via points are found while moving from a current position to a cell having a lowest move cost. As a result, the replanned part of the minimum cost path has Cell(5,3), Cell(5,4), Cell(6,5) and Cell(8,5) as via points. The function illustrated in FIG. 4A may be applied to the newly found via points to select shortest-distance via points. The mobile home appliance defines Cell(6,2), Cell(6,3), and Cell(6,4) as having an obstacle, and partially replans the path. The replanned path is expressed by a solid line in FIG. 8B. When the path is replanned, new via points are found. A minimum cost path can be recalculated based on the new via points. Since no unnecessary via point to be erased exists on the path illustrated in FIG. 8B, the path illustrated in FIG. 8B is the minimum cost path. Accordingly, the mobile home appliance moves to Cell(8,5) in the path illustrated in FIG. 8B, and thereafter, moves in an initially planned path).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Park which teaches causing the autonomous device to move from a first starting location to a second starting location; selecting a first path from the plurality of paths comprising a shortest distance between the second starting location and a first destination location; causing the autonomous device to move from the second starting location to a third starting location; and selecting a second path from the plurality of paths comprising a shortest distance between the third starting location and the first destination location with the system of Imam, as modified by MORGAN-BROWN, as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of causing the autonomous device to move from a first starting location to a second starting location; selecting a first path from the plurality of paths comprising a shortest distance between the second starting location and a first destination location; causing the autonomous device to move from the second starting location to a third starting location; and selecting a second path from the plurality of paths comprising a shortest distance between the third starting location and the first destination location and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN.
As to claim 35, Imam, as modified by MORGAN-BROWN, does not explicitly teach selecting successively shorter distances between the different starting locations and the one or more destination locations along a path between a first starting location and a first destination location; or causing the autonomous device to move along the path between the first starting location and the first destination location according to the successively shorter distances.
However, Park teaches selecting successively shorter distances between the different starting locations and the one or more destination locations along a path between a first starting location and a first destination location (see at least FIG. 2A-2C and paragraphs 38-41 regarding the path may be set by selecting a subsequent cell following a current cell. Cell(1,2), Cell(1,4), and Cell(2,2) each neighbor the cell with the start point S, i.e., Cell(1,3). Among those cells, Cell(2,2) has the lowest move cost. After moving to Cell(2,2), a cell having a lowest move cost is selected from among the cells neighboring Cell(2,2). Cell(1,2), Cell(1,1), Cell(2,1), Cell(3,1), and Cell(3,2) each neighbor Cell(2,2), and have weights 14(Cell(1,2)), 15(Cell(1,1)), 14(Cell(2,1)), 13(Cell(3,1)), and 12(Cell(3,2)). Among those cells, Cell(3,2), which has the lowest move cost, is selected. See also at least FIGS. 7A-8B and paragraphs 70-74); and causing the autonomous device to move along the path between the first starting location and the first destination location according to the successively shorter distances (see at least FIG. 2A-2C and paragraphs 38-41. See also at least FIGS. 7A-8B and paragraphs 70-74 regarding the mobile home appliance departs from Cell(1,3) and moves via Cell(2,2) and Cell(3,2)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Park which teaches selecting successively shorter distances between the different starting locations and the one or more destination locations along a path between a first starting location and a first destination location; and causing the autonomous device to move along the path between the first starting location and the first destination location according to the successively shorter distances with the system of Imam, as modified by MORGAN-BROWN, as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of selecting successively shorter distances between the different starting locations and the one or more destination locations along a path between a first starting location and a first destination location; and causing the autonomous device to move along the path between the first starting location and the first destination location according to the successively shorter distances and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN.
Claim(s) 6 is rejected under 35 U.S.C. 103 as being unpatentable over Imam et al., US 2018/0174041 A1, hereinafter referred to as Imam, in view of MORGAN-BROWN, US 2019/0316924 A1, hereinafter referred to as MORGAN-BROWN, in view of Park et al., US 2006/0149465 A1, hereinafter referred to as Park, and further in view of Chai et al., US 2021/0001897 A1, hereinafter referred to as Chai, respectively.
As to claim 6, Imam, as modified by MORGAN-BROWN and Park, does not explicitly teach wherein the one or more neural networks calculate the one or more distances in a single forward pass.
However, such matter is taught by Chai (see at least paragraph 14 regarding the system can efficiently generate the trajectory prediction output for an agent using one forward pass through a neural network model, and the trajectory prediction output can be compactly represented, e.g., by a set of probability distribution parameters).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Chai which teaches wherein the one or more neural networks calculate the one or more distances in a single forward pass with the system of Imam, as modified by MORGAN-BROWN and Park, as both systems are directed to a system and method for providing a route based on the environment information for the vehicle, and one of ordinary skill in the art would have recognized the established utility of having wherein the one or more neural networks calculate the one or more distances in a single forward pass and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN and Park.
Claim(s) 13, 14, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Imam et al., US 2018/0174041 A1, hereinafter referred to as Imam, in view of MORGAN-BROWN, US 2019/0316924 A1, hereinafter referred to as MORGAN-BROWN, in view of Park et al., US 2006/0149465 A1, hereinafter referred to as Park, and further in view of Cai et al., US 11380108 B1, hereinafter referred to as Cai, respectively.
As to claim 13, Imam, as modified by MORGAN-BROWN and Park, does not explicitly teach wherein the features are generated by one or more encoders based on a representation of the environment.
However, such matter is taught by Cai (see at least Col. 11, Line 59 – Col. 12, Line 9 regarding an encoder/decoder component 202 may receive image data that includes an image 204 captured by a sensor of an autonomous vehicle. In some examples, the encoder/decoder component 202 may include a neural network encoder (e.g., a fully connected, convolutional, recurrent, etc.) that receives the image 204 and outputs an image feature representation 206. The image feature representation 206 may include tensors associated with image features of the image 204).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Cai which teaches wherein the features are generated by one or more encoders based on a representation of the environment with the system of Imam, as modified by MORGAN-BROWN and Park, as both systems are directed to a system and method for providing a route based on the environment information, and one of ordinary skill in the art would have recognized the established utility of having wherein the features are generated by one or more encoders based on a representation of the environment and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN and Park.
As to claim 14, Imam, as modified by MORGAN-BROWN and Park, does not explicitly teach wherein the representation of the environment is an image or a point cloud.
However, such matter is taught by Cai (see at least Col. 11, Line 59 – Col. 12, Line 9 regarding an encoder/decoder component 202 may receive image data that includes an image 204 captured by a sensor of an autonomous vehicle. In some examples, the encoder/decoder component 202 may include a neural network encoder (e.g., a fully connected, convolutional, recurrent, etc.) that receives the image 204 and outputs an image feature representation 206. The image feature representation 206 may include tensors associated with image features of the image 204).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Cai which teaches wherein the representation of the environment is an image or a point cloud with the system of Imam, as modified by MORGAN-BROWN and Park, as both systems are directed to a system and method for providing a route based on the environment information, and one of ordinary skill in the art would have recognized the established utility of having wherein the representation of the environment is an image or a point cloud and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN and Park.
As to claim 19, Imam, as modified by MORGAN-BROWN and Park, does not explicitly teach wherein the representation of the environment is captured through one or more depth cameras.
However, such matter is taught by Cai (see at least Col. 27, Lines 29-39 regarding a camera may be included as a sensor on an autonomous vehicle traversing an environment. The camera may capture images of the surrounding environment as the autonomous vehicle traverses the environment).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of Cai which teaches wherein the representation of the environment is captured through one or more depth cameras with the system of Imam, as modified by MORGAN-BROWN and Park, as both systems are directed to a system and method for providing a route based on the environment information, and one of ordinary skill in the art would have recognized the established utility of having wherein the representation of the environment is captured through one or more depth cameras and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN and Park.
Claim(s) 18 is rejected under 35 U.S.C. 103 as being unpatentable over Imam et al., US 2018/0174041 A1, hereinafter referred to as Imam, in view of MORGAN-BROWN, US 2019/0316924 A1, hereinafter referred to as MORGAN-BROWN, in view of Park et al., US 2006/0149465 A1, hereinafter referred to as Park, and further in view of OGAWA et al., JP 2015080832 A, hereinafter referred to as OGAWA, respectively.
As to claim 18, Imam teaches selecting the different starting locations accessible through the step from the first location of the autonomous device (see at least FIGS. 5A-5F and paragraphs 54-63 regarding Neuron 1 may be activated following training of the SNN causing a spike to be sent from Neuron 1 to Neuron 2 at time t=0. A spike may also be sent by Neuron 1 to other neighboring neurons (e.g., Neuron 3), but due to the increased threshold potential applied across the SNN, if the synapse connecting Neuron 1 to these other neighboring neurons (e.g., Neuron 3) was not strengthened according to the STDP during training, the spike will be insufficient to meet the newly increased spiking threshold potential of the receiving neuron and that spike will not propagate further).
Imam does not explicitly teach selecting a third location of the different starting locations based at least in part on the one or more distances.
However, such matter is taught by MORGAN-BROWN (see at least FIGS. 2D-2E and paragraphs 34-35 regarding the details of each path are temporarily disregarded and the route plan is represented as a network of nodes (or graph) where each node (or vertex) is the start location, the destination location or a charging location, and each edge between nodes of the node network is the distance, time, etc. to traverse between nodes as depicted in Figure 2E. In embodiments, the determining comprises representing the start location, one or more of the electric vehicle battery charging locations in the plurality of electric vehicle battery charging locations and the destination location as nodes in a node network with paths between each node. In embodiments, the determining comprises calculating a minimum cost metric from the start location node to one or more other nodes in the node network by traversing paths in the node network from the start location node to the one or more other nodes in the node network. In embodiments, the calculating comprises accumulating the one or more cost parameters associated with each traversed path. See also at least paragraphs 80-86 regarding pre-calculating the real cost parameters between a number of source locations and target locations and then use these to train a machine learning model (such as for example a linear regression or a neural network) to predict the real cost parameters for any given source and target location. This trained model can then be installed in the vehicle and used to rapidly calculate estimates of the cost parameters between any two locations).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of MORGAN-BROWN which teaches selecting a third location of the different starting locations based at least in part on the one or more distances with the system of Imam as both systems are directed to a system and method for generating a plurality of paths based on the environment information and determining an optimal route, and one of ordinary skill in the art would have recognized the established utility of selecting a third location of the different starting locations based at least in part on the one or more distances and would have predictably applied it to improve the system of Imam.
Imam, as modified by MORGAN-BROWN and Park, does not explicitly teach calculating a size for a step of the autonomous device.
However, such matter is taught by OGAWA (see at least paragraph 63 regarding determining the robot's posture, landing position, step length, foot spacing, etc. according to the tread surface of the downstairs in order to walk down stairs).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the system of OGAWA which teaches calculating a size for a step of the autonomous device with the system of Imam, as modified by MORGAN-BROWN and Park, as both systems are directed to a system and method for providing a route based on the environment information, and one of ordinary skill in the art would have recognized the established utility of calculating a size for a step of the autonomous device and would have predictably applied it to improve the system of Imam as modified by MORGAN-BROWN and Park.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Huang et al., (US 20210347382 A1) regarding a system for generating a path for controlling an autonomous vehicle based at least in part on a static object map and/or one or more dynamic object maps.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KYLE S. PARK whose telephone number is (571)272-3151. The examiner can normally be reached Mon-Thurs 9:00AM-5:00PM.
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, Anne M ANTONUCCI can be reached at (313)446-6519. 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.
/K.S.P./Examiner, Art Unit 3666
/ANNE MARIE ANTONUCCI/Supervisory Patent Examiner, Art Unit 3666
Prosecution Insights