Teams of car-like robots push objects faster in tight spaces
MultiPush: Learning to Rearrange with Teams of Car-Like Pushers
Robotics
Summary
Moving many objects around in a small area is tricky, especially when robots that can only move like cars have to push them without bumping into each other. The authors created MultiPush, a system that helps a team of such robots plan who pushes what and when by learning from experience and using smart maps of where each robot can go. Their tests show that MultiPush speeds up the moving process and plans faster than older methods, and they even tried it with small real robots pushing 12 objects.
What this means in practice
- •For warehouse automation teams: Deploy robot teams with car-like movement to rearrange goods more quickly in confined warehouse aisles using learned scheduling and path planning.
- •For automated parking facility designers: Plan coordinated pushing tasks for robotic vehicles to rearrange cars efficiently in tight parking spaces with kinematic constraints.
Authors
Jeeho Ahn, Christoforos Mavrogiannis
Abstract
We focus on the problem of rearranging multiple objects within a constrained workspace via pushing using a team of car-like robots. While the use of multiple robots offers the potential for more efficient execution, the need for conflict resolution and the kinematic constraints arising from physics, robot design, and the workspace boundary make this problem especially challenging. Our key insight is that by exploiting the structure introduced by the car-like kinematics of the domain, we could relax the problem into an ordered assignment of Dubins curves to robots. To this end, we introduce MultiPush, a reinforcement-learning based framework that jointly determines an efficient schedule of pushing tasks and their allocation to available robots by leveraging a constraint-aware traversability graph. Across extensive simulated trials with up to 14 objects and teams of two to four robots, MultiPush reduces the makespan by up to 16% compared to the baselines while requiring up to 2.9 times faster planning time. We demonstrate MultiPush on a real-world scenario involving the rearrangement of 12 objects by two and three robots (1/10-scale racecars) in a constrained space.