Multi-Quadruped Cooperative Object Transport: Learning Decentralized Pinch-Lift-Move
Bikram Pandit, Aayam Shrestha, Alan Fern
AI summary
Problem
Coordinating multiple robots to lift and move ungraspable objects typically requires rigid mechanical coupling or centralized control, which limits flexibility and scalability.
Approach
A hierarchical policy architecture using a 'constellation reward' that aligns robot poses with the payload via point-set registration, inducing implicit synchronization without explicit communication.
Key results
- Robust transport for teams of 2 to 10 robots in simulation
- Generalization from policies trained on N=2 to larger team sizes up to N=10
- Successful manipulation of diverse geometries including boxes, logs, barrels, and couches
- Sim-to-real transfer demonstrated on physical quadruped-arm robots with lightweight objects
Why it matters
Enables scalable, communication-free cooperative transport of heavy or irregular objects in challenging environments using standard legged robots.
Abstract
We study decentralized cooperative transport us- ing teams of N-quadruped robots with arm that must pinch, lift, and move ungraspable objects through physical contact alone. Unlike prior work that relies on rigid mechanical coupling between robots and objects, we address the more challenging setting where mechanically independent robots must coordinate through contact forces alone without any communication or centralized control. To this end, we employ a hierarchical policy architecture that separates base locomotion from arm control, and propose a constellation reward formulation that unifies position and orientation tracking to enforce rigid contact behavior. The key insight is encouraging robots to behave as if rigidly connected to the object through careful reward design and training curriculum rather than explicit mechanical constraints. Our approach enables coordination through shared policy parameters and implicit synchronization cues – scaling to arbitrary team sizes without retraining. We show extensive simulation experiments to demonstrate robust transport across 2-10 robots on diverse object geometries and masses, along with sim2real transfer results on lightweight objects.