Waymo LLC

Google Inc.

Vadim Furman

Nathaniel Fairfield

Dmitri A Dolgov

Abhijit S Ogale

Joshua Seth Herbach

Andrew Hughes Chatham

Jens-Steffen Ralf Gutmann

Brandon Douglas Luders

Tobias Kunz

Michael Steven Montemerlo

Matthieu Devin

John Tisdale

Aleksandra Faust

Yu-Hsin Chen

Franklin Morley

Carlos Alberto Fuertes Pascual

Oren Eli Zamir

Rajas Moonka

Anurag Agarwal

Josh Seth Herbach

Raghava Hassan Swamy

Example systems and methods enable an autonomous vehicle to request assistance from a remote operator in certain predetermined situations. One example method includes determining a representation of an environment of an autonomous vehicle based on sensor data of the environment. Based on the representation, the method may also include identifying a situation from a predetermined set of situations for which the autonomous vehicle will request remote assistance. The method may further include sending a request for assistance to a remote assistor, the request including the representation of the environment and the identified situation. The method may additionally include receiving a response from the remote assistor indicating an autonomous operation. The method may also include causing the autonomous vehicle to perform the autonomous operation.

Remote assistance for autonomous vehicles in predetermined situations

The technology relates to identifying sensor occlusions due to the limits of the ranges of a vehicle's sensors and using this information to maneuver the vehicle. As an example, the vehicle is maneuvered along a route that includes traveling on a first roadway and crossing over a lane of a second roadway. A trajectory is identified from the lane that will cross with the route during the crossing at a first point. A second point beyond a range of the vehicle's sensors is selected. The second point corresponds to a hypothetical vehicle moving towards the route along the lane. A distance between the first point and the second point is determined. An amount of time that it would take the hypothetical vehicle to travel the distance is determined and compared to a threshold amount of time. The vehicle is maneuvered based on the comparison to complete the crossing.

Handling sensor occlusions for autonomous vehicles

A method and apparatus for controlling a first vehicle autonomously are disclosed. For instance, one or more processors may plan to maneuver the first vehicle to complete an action and predict that a second vehicle will take a particular responsive action. The first vehicle is then maneuvered towards completing the action in a way that would allow the first vehicle to cancel completing the action without causing a collision between the first vehicle and the second vehicle, and in order to indicate to the second vehicle or a driver of the second vehicle that the first vehicle is attempting to complete the action. Thereafter, when the first vehicle is determined to be able to take the action, the action is completed by controlling the first vehicle autonomously according to whether the second vehicle begins to take the particular responsive action.

Method and vehicle for responding to a lane blocking object

Aspects of the disclosure relate to identifying and adjusting speed plans for controlling a vehicle in an autonomous driving mode. In one example, an initial speed plan for controlling speed of the vehicle for a first predetermined period of time corresponding to an amount of time along a route of the vehicle is identified. Data identifying an object and characteristics of that object is received from a perception system of the vehicle. A trajectory for the object that will intersect with the route at an intersection point at particular point in time is predicted using the data. A set of constraints is generated based on at least the trajectory. The speed plan is adjusted in order to satisfy the set of constraints for a second predetermined period of time corresponding to the amount of time. The vehicle is maneuvered in the autonomous driving mode according to the adjusted speed plan.

Speed planning for autonomous vehicles

A vehicle may receive one or more images of an environment of the vehicle. The vehicle may also receive a map of the environment. The vehicle may also match at least one feature in the one or more images with corresponding one or more features in the map. The vehicle may also identify a given area in the one or more images that corresponds to a a portion of the map that is within a threshold distance to the one or more features. The vehicle may also compress the one or more images to include a lower amount of details in areas of the one or more images other than the given area. The vehicle may also provide the compressed images to a remote system, and responsively receive operation instructions from the remote system.

Image and video compression for remote vehicle assistance

Testing predictions for autonomous vehicles

Aspects of the disclosure relate to controlling a first vehicle in an autonomous driving mode. While doing so, a second vehicle may be identified. Geometry for a future trajectory of the first vehicle may be identified, and an initial allowable discomfort value may be identified. Determining a speed profile for the geometry that meets the value may be attempted by determining a discomfort value for the speed profile based on a set of factors relating to at least discomfort of a passenger of the first vehicle and discomfort of a passenger of the second vehicle. When a speed profile that meets the value cannot be determined, the value may be adjusted until a speed profile that meets the value is determined. The speed profile that meets an adjusted value is used to control the first vehicle in the autonomous driving mode.

Using discomfort for speed planning in autonomous vehicles

A vehicle may receive one or more images of an environment of the vehicle. The vehicle may also receive a map of the environment. The vehicle may also match at least one feature in the one or more images with corresponding one or more features in the map. The vehicle may also identify a given area in the one or more images that corresponds to a portion of the map that is within a threshold distance to the one or more features. The vehicle may also compress the one or more images to include a lower amount of details in areas of the one or more images other than the given area. The vehicle may also provide the compressed images to a remote system, and responsively receive operation instructions from the remote system.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for implementing an interactive autonomous vehicle agent. One of the methods includes receiving a request to generate an experience tuple for a vehicle in a particular driving context. A predicted environment observation representing a predicted environment of the autonomous vehicle after the candidate action is taken by the autonomous vehicle in an initial environment is generated, including providing an initial environment observation and the candidate action as input to a vehicle behavior model neural network trained to generate predicted environment observations. An immediate quality value is generated from a context-specific quality model that generates immediate quality values that are specific to the particular driving context. An experience tuple comprising the initial environment observation, the candidate action, and the immediate quality value is generated and used as input to a reinforcement learning system for the autonomous vehicle.

Growing community of inventors

Los Gatos, CA, United States of America

Vadim Furman