US 11,753,044 B2
Method and system for forecasting reactions of other road users in autonomous driving
Nathaniel Honka, Mountain View, CA (US)
Assigned to ARGO AI, LLC, Pittsburgh, PA (US)
Filed by Argo AI, LLC, Pittsburgh, PA (US)
Filed on Nov. 18, 2020, as Appl. No. 16/951,416.
Prior Publication US 2022/0153313 A1, May 19, 2022
Int. Cl. G05D 1/00 (2006.01); B60W 60/00 (2020.01); G05D 1/02 (2020.01); B60W 50/00 (2006.01)
CPC B60W 60/00274 (2020.02) [B60W 50/0097 (2013.01); B60W 60/0011 (2020.02); G05D 1/0088 (2013.01); G05D 1/0212 (2013.01); B60W 2420/42 (2013.01); B60W 2420/52 (2013.01); G05D 2201/0213 (2013.01)] 19 Claims
OG exemplary drawing
 
8. A method of determining a trajectory for an autonomous vehicle, the method comprising:
by an on-board computing system of the autonomous vehicle at a point in time t0:
receiving, from a presence detection subsystem of the autonomous vehicle, data indicating that a non-Ego actor is proximate to the autonomous vehicle,
generating a virtual doppelganger that is co-located with the autonomous vehicle at the point in time t0,
predicting a plurality of possible modalities for the non-Ego actor,
for each of the modalities, predicting one or more possible trajectories for the non-Ego actor,
forecasting candidate actions of the virtual doppelganger, wherein each of the candidate actions corresponds to one of the predicted possible modalities and is based on one or more of the possible trajectories of the non-Ego actor,
predicting reactive actions of the non-Ego actor, wherein each of the reactive actions is reactive to one of the candidate actions of the virtual doppelganger,
using the forecasted candidate actions of the virtual doppelganger and the predicted reactive actions of the non-Ego actor to determine candidate trajectories of the autonomous vehicle,
selecting first candidate trajectories of the candidate trajectories in which the autonomous vehicle has greatest influence on a response of the non-Ego actor,
limiting the reactive actions of the non-Ego actor based on the selected first candidate trajectories of the autonomous vehicle;
choosing one of the first candidate trajectories as a selected trajectory for the autonomous vehicle based on the limited reactive actions of the non-Ego actor, and
causing an automation subsystem of the autonomous vehicle to move the autonomous vehicle along the selected trajectory.