Space-Time Conflict Spheres for Constrained Multi-Agent Motion Planning
Session Number
Project ID: CMPS 07
Advisor(s)
Rui Chen; Carnegie Mellon University, Robotics Institute / Intelligent Control Lab
Dr. Changliu Liu; Carnegie Mellon University, Robotics Institute / Intelligent Control Lab
Discipline
Computer Science
Start Date
19-4-2023 11:55 AM
End Date
19-4-2023 12:20 PM
Abstract
1.3 million people are killed in road accidents each year, and 94% of these accidents are caused by human error. This has motivated advancements in autonomous vehicle technology, and more recently, connected autonomous vehicles (CAVs). CAVs communicate location and intention information with each other over the air, and the prospect of cooperative planning within a system of CAVs promises greater safety and efficiency in travel. CAV cooperative planning can be formulated as multi-agent motion planning (MAMP), and we propose a space-time conflict resolution approach for MAMP. We formulate the problem using a novel, flexible sphere-based trajectory representation. We then compose discrete-time procedures while evading discretization error and adhering to kinematic constraints in generated solutions. Theoretically, we prove the continuous-time feasibility and formulation-space completeness of our algorithm. Experimentally, we demonstrate that our algorithm matches the performance of the current state of the art with respect to runtime and solution quality, while expanding upon the abilities of current work through accommodation for both static and dynamic obstacles. We evaluate our algorithm in various unsignalized traffic intersection scenarios using CARLA, an open-source vehicle simulator. Results show significant success rate improvement in spatially constrained settings and performance that scales well among increasingly complex scenarios.
Space-Time Conflict Spheres for Constrained Multi-Agent Motion Planning
1.3 million people are killed in road accidents each year, and 94% of these accidents are caused by human error. This has motivated advancements in autonomous vehicle technology, and more recently, connected autonomous vehicles (CAVs). CAVs communicate location and intention information with each other over the air, and the prospect of cooperative planning within a system of CAVs promises greater safety and efficiency in travel. CAV cooperative planning can be formulated as multi-agent motion planning (MAMP), and we propose a space-time conflict resolution approach for MAMP. We formulate the problem using a novel, flexible sphere-based trajectory representation. We then compose discrete-time procedures while evading discretization error and adhering to kinematic constraints in generated solutions. Theoretically, we prove the continuous-time feasibility and formulation-space completeness of our algorithm. Experimentally, we demonstrate that our algorithm matches the performance of the current state of the art with respect to runtime and solution quality, while expanding upon the abilities of current work through accommodation for both static and dynamic obstacles. We evaluate our algorithm in various unsignalized traffic intersection scenarios using CARLA, an open-source vehicle simulator. Results show significant success rate improvement in spatially constrained settings and performance that scales well among increasingly complex scenarios.