Motion planning runs up to 10 times faster with vectorized reparameterization
ReVAMP: Vector-Accelerated Motion Planning for Kinematically-Constrained Systems via Reparameterization
Robotics
Summary
Robots often have to move while following strict rules about how their arms or tools can be positioned. These rules make the planning very tricky because only a tiny set of movements are allowed. The authors found a new way to re-describe the robot’s possible movements so they can quickly find good paths by doing many calculations at once. Their method lets robots calculate complex movement plans much faster than before, which could make robot tasks more efficient and flexible.
What this means in practice
- •For industrial robot programmers: Generate feasible robot arm motions quickly under complex constraints to improve manufacturing flexibility.
- •For autonomous vehicle software teams: Plan paths efficiently for vehicles with mechanical or environmental constraints using accelerated analytic reparameterization.
Authors
Shrutheesh R. Iyer, Thomas Cohn, Zachary Kingston
Abstract
Robots often must satisfy one or more constraints during motion planning for real-world tasks. When such constraints reduce the valid configuration space to a measure-zero subset, sampling based planning algorithms require modifications to draw feasible samples. For many common end-effector constraints, parameterizations built on inverse kinematics (IK) provide an alternate formulation where the constraints are satisfied by construction, allowing directly sampling the feasible set. Despite their elegant approach, parameterized planners have remained slower than vector-accelerated implementations of projection-based approaches, leaving their performance ceiling an open question. We explore a new axis of vectorization built upon reparameterizing the planning space through analytic IK. This approach addresses existing inefficiencies in vectorized projection-based planners and exposes new opportunities for parallelism within the planner. We show that the planner can synthesize plans in microseconds to milliseconds for high dimensional systems (up to 20 dimensions), with complex constraints, up to 10x faster than the current state-of-the-art. Furthermore, we demonstrate how such planning speeds open up avenues for restructuring sequential manipulation pipelines.