Fast motion planning and control in the era of parallelism

Please register here: https://forms.gle/tERfG19MbS9CiA5r7 if you are interested in joining the workshop remotely.

A workshop on recent advances in efficient motion planning and control via parallelism

Motion planning and control are at the core of every autonomous robot system, ranging from mobile robot navigation, to manipulation and locomotion. A common assumption in the field is that motion planning provides a reference trajectory at a slow speed while a controller generates a control input at a higher speed to realize the trajectory. Recent advances in motion planning via CPU- and GPU-based parallelism have reduced the planning time to a few milliseconds, and in many cases, even microseconds, blurring the traditional line between motion planning and control and also unfolding a promising research area in robot automation. This workshop aims to discuss the latest breakthroughs and brainstorm ideas in motion planning and control via parallelism techniques and re-examine their roles in many different aspects of robot autonomy, including but not limited to: planning efficiency, optimality, reactiveness, perception-compatibility, practicality and robot safety. 

We invite paper submissions on various aspects of motion planning and control via parallelization, including but not limited to the following topics:

• How can parallelism improve sampling-based, search-based, optimization-based and learning-based motion planning?

• How can parallelism improve robot control, e.g., model predictive control (MPC) techniques?

• How can one account for nonlinear constraints such as robot dynamics, contact dynamics, state space manifolds, etc. with parallelization techniques?

• What environment representations are compatible with parallelized motion planning and control?

• What are the new roles of planning and control in robot autonomy, as parallelized motion planning can achieve microseconds to milliseconds planning time?

As a non-archival workshop, we also welcome recent published or submitted results on the relevant topics. The submissions will be reviewed, and if accepted, will be posted on the workshop website and presented at our poster session.

Date: June 21, 2025

Location: University of Southern California

Submission Portal: OpenReview

Paper Format: 2-4 pages (excluding references), single-blind RSS format 

Submission Deadline: May 28, 2025, 11:59pm PDT

Notification of Acceptance: June 4, 2025