# US 2020/0401149 A1 · Corner case detection and collection for a path planning system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2020/0401149 A1 - Document ID: US20200401149A1 - Status: Published application - Kind: A1 - Application number: 16/448,070 - Filed: 2019-06-21 - Published: 2020-12-24 - Earliest claimed: 2019-06-21 - Assignees: Volkswagen Aktiengesellschaft; Audi AG; Porsche AG - Applicants: Volkswagen AG; Audi AG; Porsche AG - Inventors: Xinwei GONG; Sai Prasad NOOKA - Web page: https://patentgrep.com/patent/US20200401149A1 ## CPC - G05D 1/0221: {involving a learning process} - G06Q 10/00: Administration; Management - G06F 11/0739: {in a data processing system embedded in automotive or aircraft systems} - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/09: Supervised learning - G06Q 10/047: Optimisation of routes or paths, e.g. travelling salesman problem - G06Q 50/40: Business processes related to the transportation industry (shipping G06Q10/083) - G05D: SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Disclosed embodiments provide systems and methods for detecting and storing information describing errors in a predictive path planning functionality of a neural network. ## Claims ### Claim 1 Transportation vehicle equipment for detecting errors in a path predicting neural network, the equipment comprising: one or more sensors disposed on a transportation vehicle for acquiring data describing a roadway and a reference path taken by the transportation vehicle; a processor and memory; at least one neural network for predicting one or more predicted paths for the transportation vehicle; and means for evaluating the one or more predicted paths to detect one or more errors in the predicting operation wherein the one or more errors include an indication that the one or more predicted paths of the path planning system deviate from the reference path. ### Claim 2 (depends on 1) The transportation vehicle equipment of claim 1, wherein the means for evaluating compares the one or more predicted paths to the reference path. ### Claim 3 (depends on 2) The transportation vehicle equipment of claim 2, wherein the means for evaluating detects deviation of the one or more predicted paths from the reference path. ### Claim 4 (depends on 3) The transportation vehicle equipment of claim 3, further comprising memory for storing data describing the one or more predicted paths and the reference path. ### Claim 5 (depends on 4) The transportation vehicle equipment of claim 4, further comprising memory for storing data from the one or more sensors continuously in a ring buffer. ### Claim 6 (depends on 5) The transportation vehicle equipment of claim 5, further comprising memory for transmitting data from the ring buffer to a semi-permanent memory in response to detection of the one or more errors in the predicting operation. ### Claim 7 (depends on 6) The transportation vehicle equipment of claim 6, wherein the data transmitted from the ring buffer to the semi-permanent memory includes a first portion of data corresponding to a period of time before the detected error and a portion of data corresponding to a period of time after the detected error. ### Claim 8 (depends on 1) The transportation vehicle equipment of claim 1, wherein the means for evaluating determines a smoothness metric from the one or more predicted paths. ### Claim 9 (depends on 8) The transportation vehicle equipment of claim 8, wherein the one or more predicted paths comprises a plurality of predicted paths, and wherein the smoothness metric is based in part on a comparison of the plurality of predicted paths to one another. ### Claim 10 (depends on 1) The transportation vehicle equipment of claim 1, wherein the one or more predicted path comprises a plurality of predicted paths, and wherein the plurality of predicted paths are generated by more than one neural network. ### Claim 11 (depends on 1) The transportation vehicle equipment of claim 1, wherein the one or more predicted path comprises a plurality of predicted paths, and wherein the plurality of predicted paths are generated by one neural network according to dropout sampling. ### Claim 12 (depends on 10) The transportation vehicle equipment of claim 10, wherein the means for evaluating compares the plurality of predictive paths and detects variance between the plurality of predictive paths to measure an uncertainty value. ### Claim 13 (depends on 11) The transportation vehicle equipment of claim 11, wherein the means for evaluating compares the plurality of predictive paths and detects variance between the plurality of predictive paths to measure an uncertainty value. ### Claim 14 A method of error detection for a path predicting neural network, the method comprising: analyzing data describing a roadway acquired by one or more sensors disposed on a transportation vehicle; predicting with at least one neural network one or more predicted paths for the transportation vehicle; acquiring data describing a reference path taken by the transportation vehicle; and evaluating the one or more predicted paths to detect one or more errors in the predicting operation wherein the one or more errors include an indication that the one or more predicted paths of the path planning system deviate from the reference path. ### Claim 15 (depends on 14) The method of error detection for the path predicting neural network of claim 14, further comprising comparing the one or more predicted paths to the reference path. ### Claim 16 (depends on 15) The method of error detection for the path predicting neural network of claim 15, further comprising detecting deviation of the one or more predicted paths from the reference path. ### Claim 17 (depends on 16) The method of error detection for the path predicting neural network of claim 16, further comprising storing data describing the one or more predicted paths and the reference path. ### Claim 18 (depends on 17) The method of error detection for the path predicting neural network of claim 17, further comprising storing data from the one or more sensor continuously in a ring buffer. ### Claim 19 (depends on 18) The method of error detection for the path predicting neural network of claim 18, further comprising transmitting data from the ring buffer to a semi-permanent memory in response to detection of the one or more error in the predicting operation. ### Claim 20 (depends on 19) The method of error detection for the path predicting neural network of claim 19, wherein the data transmitted from the ring buffer to the semi-permanent memory includes a first portion of data corresponding to a period of time before the detected error and a portion of data corresponding to a period of time after the detected error. ### Claim 21 (depends on 14) The method of error detection for the path predicting neural network of claim 14, further comprising determining a smoothness metric from the one or more predicted paths. ### Claim 22 (depends on 21) The method of error detection for the path predicting neural network of claim 21, wherein the one or more predicted path comprises a plurality of predicted paths, and wherein the smoothness metric is based at least in part on a comparison of the plurality of predicted paths to one another. ### Claim 23 (depends on 14) The method of error detection for the path predicting neural network of claim 14, wherein the one or more predicted path comprises a plurality of predicted paths, and wherein the plurality of predicted paths are generated by more than one neural network. ### Claim 24 (depends on 14) The method of error detection for the path predicting neural network of claim 14, wherein the one or more predicted path comprises a plurality of predicted paths, and wherein the plurality of predicted paths are generated by the neural network. ### Claim 25 (depends on 24) The method of error detection for the path predicting neural network of claim 24, wherein the plurality of predicted paths are generated by the neural network according to dropout sampling. ### Claim 26 (depends on 23) The method of error detection for the path predicting neural network of claim 23, further comprising comparing the plurality of predictive paths, and detecting variance between the plurality of predictive paths to measure an uncertainty value. ### Claim 27 (depends on 24) The method of error detection for the path predicting neural network of claim 24, further comprising comparing the plurality of predictive paths, and detecting variance between the plurality of predictive paths to measure an uncertainty value. ### Claim 28 A computing system for planning a path of a vehicle, comprising: at least one memory configured to receive sensor data indicative of a reference path for the vehicle; at least one processor communicatively coupled to the at least one memory, wherein the at least one processor is configured to execute program instructions to cause the system to perform the steps comprising: predicting one or more predicted paths for the vehicle by at least one neural network; detecting a deviation from the one or more predicted paths; determining a corner case based on the reference path and the detected deviation; storing the corner case; and training the at least one neural network with the stored corner case.