US 11,697,429 B2
System, method, and computer program product for topological planning in autonomous driving using bounds representations
Neal Andrew Seegmiller, Pittsburgh, PA (US); Patrick Stirling Barone, San Francisco, CA (US); and Arek Viko Sredzki, San Mateo, CA (US)
Assigned to Argo AI, LLC, Pittsburgh, PA (US)
Filed by Argo AI, LLC, Pittsburgh, PA (US)
Filed on Feb. 10, 2021, as Appl. No. 17/172,530.
Prior Publication US 2022/0250641 A1, Aug. 11, 2022
Int. Cl. B60W 60/00 (2020.01); G01C 21/34 (2006.01)
CPC B60W 60/001 (2020.02) [G01C 21/3453 (2013.01); B60W 2556/50 (2020.02)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method of maneuvering an autonomous vehicle traversing a route on a roadway, comprising:
receiving, by the autonomous vehicle, map data and sensor data associated with a plurality of constraints in a geographic area of the autonomous vehicle;
expanding, by a computing system of the autonomous vehicle, a topological tree by adding a plurality of nodes to represent a plurality of actions associated with the plurality of constraints by:
generating, by the computing system, a bound based on a constraint of the plurality of constraints, each constraint of the plurality of constraints specifying a semantic action for traversing an object in the roadway in the geographic area, and the bound identifying limits of a spatial area with respect to an action for navigating the autonomous vehicle relative to the constraint;
storing, by a central database management system of the computing system, bound data, including a bound identifier associated with the bound, in a central bound storage; and
linking, by the computing system, the bound and one or more additional bounds sharing a common constraint in the plurality of nodes of the topological tree to the central bound storage, based on the bound identifier, wherein the bound is initially linked as an active bound, or alternatively, as an inactive bound after determining it is not a most restrictive bound at any sample index; and
controlling, by the computing system, the autonomous vehicle based on the topological tree, to navigate the plurality of constraints.