US 11,808,590 B2
Autonomous navigation in a cluttered environment
Daniela Rus, Weston, MA (US); Sertac Karaman, Cambridge, MA (US); Wilko Schwarting, Cambridge, MA (US); Anshula Gandhi, Cambridge, MA (US); Cristian-Ioan Vasile, Cambridge, MA (US); and Alyssa Pierson, Somerville, MA (US)
Assigned to Massachusetts Institute of Technology, Cambridge, MA (US)
Filed by Massachusetts Institute of Technology, Cambridge, MA (US)
Filed on Jan. 13, 2020, as Appl. No. 16/741,039.
Claims priority of provisional application 62/791,383, filed on Jan. 11, 2019.
Prior Publication US 2020/0225053 A1, Jul. 16, 2020
Int. Cl. G01C 21/34 (2006.01); G01C 21/20 (2006.01); B60W 60/00 (2020.01)
CPC G01C 21/3461 (2013.01) [B60W 60/0015 (2020.02); G01C 21/20 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A method for vehicle navigation comprising planning a route from a current location of a vehicle to a destination location in an environment, the planning including:
acquiring sensor data characterizing locations of objects in the environment, at least some of the objects being non-permanent moving objects;
determining an occupancy density ρ(q) over locations in the environment from the sensor data, where q represents a location in a two-dimensional or three-dimensional physical environment of the vehicle;
determining a velocity field over locations in the environment from the occupancy density and at least one prior occupancy density determined from prior sensor data, wherein determining the velocity field comprises determining an occupancy density flow by processing the occupancy density and the prior occupancy density using optical flow techniques;
determining a risk map for locations in the environment from the occupancy density and the velocity field;
determining a path toward the destination location using the risk map in a route planning procedure; and
causing the vehicle to traverse at least an initial portion of the path.