US 12,441,368 B2
Vehicle action selection based on simulated states
Caglayan Dicle, Melrose, MA (US); Benjamin Riviere, Pasadena, CA (US); Eric Wolff, Zephyr, NV (US); and Dmitry S. Yershov, Medford, MA (US)
Assigned to Motional AD LLC, Boston, MA (US)
Filed by Motional AD LLC, Boston, MA (US)
Filed on Apr. 6, 2022, as Appl. No. 17/658,243.
Prior Publication US 2023/0322266 A1, Oct. 12, 2023
Int. Cl. B60W 60/00 (2020.01); B60W 50/00 (2006.01); G05B 13/02 (2006.01)
CPC B60W 60/0027 (2020.02) [B60W 50/0097 (2013.01); B60W 60/0013 (2020.02); B60W 60/0015 (2020.02); G05B 13/0265 (2013.01); B60W 2050/0022 (2013.01); B60W 2050/0052 (2013.01); B60W 2420/403 (2013.01); B60W 2420/408 (2024.01); B60W 2556/40 (2020.02); B60W 2720/10 (2013.01); B60W 2720/106 (2013.01); B60W 2720/24 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method, comprising:
obtaining scene data associated with a scene of a vehicle;
determining a scene state of the scene based on the scene data;
determining a plurality of available actions for the vehicle using the scene state, wherein a particular available action of the plurality of available actions corresponds to a particular instruction to control the vehicle;
determining a plurality of simulated scene states using the plurality of available actions, wherein a particular simulated scene state of the plurality of simulated scene states corresponds to a simulation of the scene state resulting from the particular action of the plurality of available actions;
generating a plurality of simulated trajectories for the vehicle using the plurality of simulated scene states, wherein a particular simulated trajectory of the plurality of simulated trajectories begins at the particular simulated scene state, wherein generating the particular simulated trajectory of the plurality of simulated trajectories includes simulating concurrent actions taken by a plurality of objects in the scene of the vehicle;
evaluating the plurality of simulated trajectories;
selecting an action from the plurality of available actions corresponding to a simulated scene state of the plurality of simulated scene states based on the evaluating the plurality of simulated trajectories; and
causing the vehicle to execute the action.