Zoox, Inc.

Joona Markus Petteri Kiiski

Marin Kobilarov

Timothy Caldwell

Jacob Lee Askeland

Robert Edward Somers

Vasumathi Raman

Christopher Paxton

Joseph Funke

Ashutosh Gajanan Rege

Gary Linscott

Lichao Ma

Lakshay Garg

Ravi Verma Gogna

Caili Li

Robert Jonathan Crane

The present disclosure involves systems and methods for detecting road blockages and generating alternative routes. In some cases, a system detects, based at least in part on sensor data associated with an autonomous vehicle, a portion of an environment that impedes a planned path of the autonomous vehicle. The system determines a semantic classification associated with the portion of the environment and transmits a re-routing request comprising the semantic classification to one or more remote computing devices. The system receives an instruction associated with navigating the autonomous vehicle around the portion of the environment from the remote computing devices, where the instruction includes an alternative route determined from a plurality of alternative routes. The system further controls the autonomous vehicle to navigate around the portion of the environment based at least in part on the instruction.

Techniques for detecting road blockages and generating alternative routes

Techniques for determining right of way through an intersection are discussed herein. Routes through the intersection may be associated with respective priorities. The route associated with an inbound lane devoid of yield or stop markers may be determined as being associated with the highest priority. The hierarchy of the other priorities may be organized based on whether the number of times routes associated with each respective priority intersects the route associated with the highest priority. The routes and the priorities are saved in a data structure, and the data structure is transmitted to an autonomous vehicle for controlling the autonomous vehicle through the intersection.

Determining right of way

A vehicle computing system may implement techniques to determine relevance of objects detected in an environment to a vehicle operating in the environment. The techniques may include determining locations and trajectories associated with a detected object and generating simulations of movement (e.g., estimated states) of the detected object and the vehicle over a period of time. The vehicle computing system may compare locations, distances, and/or trajectories of the detected object and the vehicle in an estimated state to determine whether the detected object is relevant to the vehicle performing the action (e.g., impacts the vehicle's ability to safely perform the action). The relevance determination may be based in part on a semantic classification of the detected object. Based on a determination that the detected object is irrelevant to the vehicle in an estimated state, the vehicle computing device may disregard the detected object in future estimated states and/or future vehicle control planning considerations.

Dynamic object relevance determination

Trajectory generation and/or execution architecture is described. In an example, a first signal can be determined at a first frequency, wherein the first signal comprises information associated with causing the system to move to a location. Further, a second signal can be determined at a second frequency different from the first frequency and based at least in part on the first signal. A system can be controlled to move to the location, based at least in part on the second signal.

Trajectory generation and execution architecture

Techniques and methods for determining a distance between a point within an environment and a reference line are discussed herein. For instance, a vehicle may be navigating. While navigating, the vehicle may receive a reference line that represents a road segment and determine various regions relative to the reference line. Additionally, the vehicle may generate sensor data representing the environment and identify an object using the sensor data. The vehicle may then determine that a location of the object corresponds to a region from the regions. Based on the region, the vehicle may determine a rule for identifying the distance between the vehicle and the reference line. The vehicle may then determine the distance using the rule, the location of the object, and the reference line. Additionally, the vehicle may determine an action for the vehicle to perform that is based on the distance.

Techniques for determining a distance between a point and a spiral line segment

Controlling autonomous vehicles requires accurate information about how entities behave in an environment with respect to one another. In an example, a data structure is used to associate entities in an environment of the vehicle with the vehicle and/or with each other. For example, associations between entities can be based on locations of the entities relative to each other and/or relative to segments of a drivable surface. Information about associations between the vehicle and/or entities in the environment may be used to identify entities that may influence travel of the vehicle and/or other entities and, in some implementations, vehicle controls can be generated based on these relevant entities.

Vehicle control using directed graphs

Techniques for determining a trajectory for an autonomous vehicle are described herein. In general, determining a route can include utilizing a search algorithm such as Monte Carlo Tree Search (MCTS) to search for possible trajectories, while using temporal logic formulas, such as Linear Temporal Logic (LTL), to validate or reject the possible trajectories. Trajectories can be selected based on various costs and constraints optimized for performance. Determining a trajectory can include determining a current state of the autonomous vehicle, which can include determining static and dynamic symbols in an environment. A context of an environment can be populated with the symbols, features, predicates, and LTL formula. Rabin automata can be based on the LTL formula, and the automata can be used to evaluate various candidate trajectories. Nodes of the MCTS can be generated and actions can be explored based on machine learning implemented as, for example, a deep neural network.

Trajectory generation using temporal logic and tree search

This application describes locating objects in an environment, such as a mapped region. In an example, a location of an entity on a drivable surface can include querying an axis-aligned bounding box (AABB) tree to determine a segment of the drivable surface including the location. In some examples, the location can be further determined within the segment of the drivable surface. For instance, the segment of the drivable surface may be partitioned into a plurality of polygons, and the polygons can be searched to determine a lane segment in the road segment associated with the entity.

Locating entities in a mapped environment

Techniques for generating and executing trajectories to guide autonomous vehicles are described. In an example, a first computer system associated with an autonomous vehicle can generate, at a first operational frequency, a route to guide the autonomous vehicle from a current location to a target location. The first computer system can further determine, at a second operational frequency, an instruction for guiding the autonomous vehicle along the route and can generate, at a third operational frequency, a trajectory based at least partly on the instruction and real-time processed sensor data. A second computer system that is associated with the autonomous vehicle and is in communication with the first computer system can execute, at a fourth operational frequency, the trajectory to cause the autonomous vehicle to travel along the route. The separation of the first computer system and the second computer system can provide enhanced safety, redundancy, and optimization.

Growing community of inventors

Cupertino, CA, United States of America

Joona Markus Petteri Kiiski