New method helps groups of robots coordinate movement
A protocol called WarpWeaver uses negotiation to help robots reach near-optimal agreements.
Researchers developed WarpWeaver to coordinate a multi-robot team. This decentralized, real-time planner approaches the joint optimum as an anytime process.
The method extends a single-robot framework that searches dispersive trees of forward-propagated commands. Inter-robot coupling is replaced by local constraints. These constraints allow locally grown dispersive trees to contain a near-optimal agreement.
Robots use decentralized column generation to pursue this structure. They negotiate rents on the shared space to guide the search of each robot's tree. This negotiation settles on an agreement whose near-optimality is certified at quiescence.
WarpWeaver implements this negotiation as an asynchronous real-time loop over a receding horizon. It shows effective closure to quiesced agreements even when real-time deadlines cut the negotiation short.
In simulation, WarpWeaver reduces total travel time by up to 22% against centralized offline planners on the Dynobench suite. It reduces travel time by up to 37% against a decentralized real-time baseline. It nearly doubles the success rate of a task-specific planner on graph connectivity.
Hardware experiments on six robots planning onboard demonstrate real-world implementability.
Key facts
- WarpWeaver reduces total travel time by up to 22% against centralized offline planners on the Dynobench suite.
- The protocol reduces travel time by up to 37% against a decentralized real-time baseline.
- Hardware experiments were performed on six robots planning onboard.