Motional Ad LLC

Nutonomy Inc.

Massachusetts Institute of Technology

Nanyang Technological University

Tichakorn Wongpiromsarn

Scott D Pendleton

Oscar Olof Beijbom

Shih-Yuan Liu

Sourabh Vora

Radboud Duintjer Tebbens

Calin Belta

Anne Collin

Daniele De Francesco

Karl D Iagnemma

Eric Wolff

Harshavardhan Ravichandran

Emilio Frazzoli

Dmytro S Yershov

Bassam Helou

Jeong Hwan Jeon

Hsun-Hsien Chang

Kostyantyn Slutskyy

Yu Wang

Amitai Bin-Nun

Tawit Uthaicharoenpong

Danwei Wang

Techniques are provided for operation of a vehicle using multiple motion constraints. The techniques include identifying an object using one or more processors of a vehicle. The vehicle has a likelihood of collision with the object that is greater than a threshold. The processors generate multiple motion constraints for operating the vehicle. At least one motion constraint includes a minimum speed of the vehicle greater than zero to avoid a collision of the vehicle with the object. The processors identify one or more motion constraints for operating the vehicle to avoid a collision of the vehicle with the object. The processors operate the vehicle in accordance with the identified motion constraints.

Operation of a vehicle using multiple motion constraints

Techniques are provided for autonomous vehicle operation using linear temporal logic. The techniques include using one or more processors of a vehicle to store a linear temporal logic expression defining an operating constraint for operating the vehicle. The vehicle is located at a first spatiotemporal location. The one or more processors are used to receive a second spatiotemporal location for the vehicle. The one or more processors are used to identify a motion segment for operating the vehicle from the first spatiotemporal location to the second spatiotemporal location. The one or more processors are used to determine a value of the linear temporal logic expression based on the motion segment. The one or more processors are used to generate an operational metric for operating the vehicle in accordance with the motion segment based on the determined value of the linear temporal logic expression.

Autonomous vehicle operation using linear temporal logic

Techniques for operation of a vehicle using machine learning with motion planning include storing, using one or more processors of a vehicle located within an environment, a plurality of constraints for operating the vehicle within the environment. One or more sensors of the vehicle receive sensor data describing the environment. The one or more processors extract a feature vector from the stored plurality of constraints and the received sensor data. The feature vector includes a first feature describing an object located within the environment. A machine learning circuit of the vehicle is used to generate a first motion segment based on the feature vector. A number of violations of the stored plurality of constraints is below a threshold. The one or more processors operate the vehicle in accordance with the generated first motion segment.

Operation of a vehicle using motion planning with machine learning

Among other things, a system provides speed behavior planning for vehicles with autonomous driving capabilities.

Speed behavior planning for vehicles

Enclosed are embodiments for scoring one or more trajectories of a vehicle through a given traffic scenario using a machine learning model that predicts reasonableness scores for the trajectories. In an embodiment, human annotators, referred to as a 'reasonable crowd,' are presented with renderings of two or more vehicle trajectories traversing through the same or different traffic scenarios. The annotators are asked to indicate their preference for one trajectory over the other(s). Inputs collected from the human annotators are used to train the machine learning model to predict reasonableness scores for one or more trajectories for a given traffic scenario. These predicted trajectories can be used to rank trajectories generated by a route planner based on their scores, compare AV software stacks, or used by any other application that could benefit from a machine learning model that scores vehicle trajectories.

Scoring autonomous vehicle trajectories using reasonable crowd data

Enclosed are embodiments for scenario-based behavior specification and validation. In an embodiment, a method comprises: obtaining, using at least one processor, at least one trajectory associated with a driving scenario for an autonomous vehicle system; evaluating, using the at least one processor and at least one rulebook, the at least one trajectories to determine whether the at least one trajectory violates at least one rule in the at least one rulebook, wherein each rule in the rulebook is associated with at least one violation metric that is used to determine a degree to which the rule was satisfied or violated; determining, using the at least one processor and the at least one violation metric, a score for the at least one trajectory; and sending, using the at least one processor, the score to at least one of a software module in a software stack of the autonomous vehicle system, a simulation of the autonomous vehicle system or as a report or in a visual presented through a user interface of a cloud-based platform.

Scenario-based behavior specification and validation

Among other things, a vehicle is caused to drive autonomously through a road network toward a defined goal position at which a stopped activity is to occur. A set of acceptable stopping places is identified in the vicinity of the goal position. The identifying of the set includes: identifying a proximity region in the vicinity of the goal position, identifying a goal region within the proximity region, discretizing the goal region to identify acceptable stopping places given characteristics of the vehicle and of the stopped activity analyzing current information about potential stopping places in the goal region in the vicinity of the goal position to make a choice of a currently selected stopping place that is acceptable and feasible. Current information is analyzed about the potential stopping places, to make a choice of a currently selected stopping place that is acceptable and feasible the vehicle is caused to drive autonomously toward the currently selected stopping place. The activities are repeated until the vehicle stops at a currently selected stopping place.

Identifying a stopping place for an autonomous vehicle

Among other things, a vehicle is caused to drive autonomously through a road network toward a defined goal position. Current information is analyzed about potential stopping places in the vicinity of the goal position, to make a choice of a currently selected stopping place that is acceptable and feasible. The vehicle is caused to drive autonomously toward the currently selected stopping place. The activities are repeated until the vehicle stops at a currently selected stopping place.

A distributed traffic signal control method is provided for a directed network comprising a plurality of junctions, each junction having a plurality of links connected thereto, the links comprising one or more upstream links and one or more downstream links, the method comprising: activating one of a plurality of phases of the junction for a predetermined time period which maximizes the directed network throughput based on current differential traffic backlogs between said one or more upstream links and said one or more downstream links, each phase providing a unique combination of traffic signals at the junction for guiding traffic from the upstream link(s) to the downstream link(s). There is also provided a corresponding traffic signal controller, a traffic control system comprising the traffic signal controller, and a computer readable medium having stored therein computer executable codes for instructing a computer processor to execute the distributed traffic signal control method.

Growing community of inventors

Singapore, Singapore

Tichakorn Wongpiromsarn