Motional Ad LLC

Hans Andersen

Ning Wu

Scott D Pendleton

Francesco Seccamonte

Sammy Omari

Bence Cserna

Zachary Thomas Batts

Juraj Kabzan

Kevin C Gall

Qian Wang

Marc Dominik Heim

Yiming Zhao

Lixun Lin

Boaz Cornelis Floor

Jesse Adam Miller

Xiyuan Liu

Provided are methods and systems for semantic behavior filtering for prediction improvement. A method for operating an autonomous vehicle, is provided. The method includes obtaining, by at least one processor, semantic image data associated with an environment where an autonomous vehicle is operating. The method includes determining, by the at least one processor, a set of agents in the environment based on the semantic image data. The method includes determining a set of predicted actions for at least one agent of the set of agents. The method includes determining, from the set of predicted actions, a set of secondary predicted actions for the at least one primary agent using semantic data. The method includes determining, from the set of predicted actions, a set of primary predicted actions other than secondary predicted actions The method includes generating a path for the autonomous vehicle based on the set of primary predicted actions.

Path generation based on predicted actions

This disclosure describes an autonomous vehicle configured to obtain sensor data associated with objects proximate a projected route of the autonomous vehicle, determine static constraints that limit a trajectory of the autonomous vehicle along the projected route based on non-temporal risks associated with a first subset of the f objects, predict a position and speed of the autonomous vehicle as a function of time along the projected route based on the static constraints, identify temporal risks associated with a second subset of the objects based on the predicted position and speed of the autonomous vehicle, determine dynamic constraints that further limit the trajectory of the autonomous vehicle along the projected route to help the autonomous vehicle avoid the temporal risks associated with the second subset of the objects, and adjust the trajectory of the autonomous vehicle in accordance with the static constraints and the dynamic constraints.

Planning with dynamic state a trajectory of an autonomous vehicle

Among other things, we describe techniques for electric power steering torque compensation. Techniques are provided for a method implemented by a computer, e.g., a computer onboard an autonomous vehicle. A planning circuit onboard the vehicle and connected to an EPS of the vehicle determines a compensatory torque signal to modify an actual steering angle of a steering wheel of the vehicle to match an expected steering angle of the steering wheel. The planning circuit transmits the compensatory torque signal to a control circuit that controls the steering angle of the steering wheel. The EPS modifies the actual steering angle based on the compensatory torque signal resulting in a modified steering angle. The control circuit operates the vehicle based on the modified steering angle.

Electric power steering torque compensation

The subject matter described in this specification is generally directed control architectures for an autonomous vehicle. In one example, a reference trajectory, a set of lateral constraints, and a set of speed constraints are received using a control circuit. The control circuit determines a set of steering commands based at least in part on the reference trajectory and the set of lateral constraints and a set of speed commands based at least in part on the set of speed constraints. The vehicle is navigated, using the control circuit, according to the set of steering commands and the set of speed commands.

Control architectures for autonomous vehicles

Growing community of inventors

Singapore, Singapore

Hans Andersen