US 11,755,015 B2
Scoring autonomous vehicle trajectories using reasonable crowd data
Oscar Olof Beijbom, Santa Monica, CA (US); Bassam Helou, Santa Monica, CA (US); Radboud Duintjer Tebbens, Winchester, MA (US); Calin Belta, Sherborn, MA (US); Anne Collin, Cambridge, MA (US); and Tichakorn Wongpiromsarn, Ames, IA (US)
Assigned to Motional AD LLC, Boston, MA (US)
Filed by Motional AD LLC, Boston, MA (US)
Filed on Dec. 20, 2021, as Appl. No. 17/555,941.
Application 17/555,941 is a continuation of application No. 17/336,685, filed on Jun. 2, 2021, granted, now 11,203,362.
Application 17/336,685 is a continuation of application No. 17/009,656, filed on Sep. 1, 2020.
Prior Publication US 2022/0204033 A1, Jun. 30, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/08 (2023.01); G05D 1/00 (2006.01); B60W 60/00 (2020.01); G01C 21/36 (2006.01); G01C 21/34 (2006.01); G05D 1/02 (2020.01); G06V 20/40 (2022.01); G06V 20/56 (2022.01); G06F 18/21 (2023.01); G06F 18/214 (2023.01); G06F 18/2415 (2023.01); G06N 5/01 (2023.01); G06V 10/82 (2022.01); G06V 20/54 (2022.01)
CPC G05D 1/0088 (2013.01) [B60W 60/0011 (2020.02); G01C 21/3407 (2013.01); G01C 21/3691 (2013.01); G05D 1/0212 (2013.01); G06F 18/217 (2023.01); G06F 18/2148 (2023.01); G06F 18/2415 (2023.01); G06N 3/08 (2013.01); G06N 5/01 (2023.01); G06V 10/82 (2022.01); G06V 20/41 (2022.01); G06V 20/54 (2022.01); G06V 20/56 (2022.01); B60W 2552/00 (2020.02)] 21 Claims
OG exemplary drawing
 
15. A system, comprising:
at least one processor, and
at least one non-transitory storage media storing instructions that, when executed by the at least one processor, cause the at least one processor to:
augment a route planner of a first vehicle using predicted reasonableness scores for trajectories, wherein reasonableness scores are predicted by a trained machine learning model with parameters determined using a loss function that penalizes predictions of reasonableness scores that violate a rulebook structure;
plan trajectories in an environment using the augmented route planner;
identify at least one trajectory with inadequate performance by the augmented route planner based on a first metric associated with the predicted reasonableness scores; and
change parameters of the augmented route planner in response to the identified at least one trajectory.