Overview
This project addresses how robots coordinate in shared spaces through dynamic planning and task negotiation. By developing sophisticated spatial planning algorithms and communication protocols, teams of robots can work together efficiently in complex environments, enabling scalable teamwork in real-world assembly and manufacturing scenarios.
Objectives
- Develop dynamic spatial planning algorithms for multi-robot systems
- Create efficient task negotiation and allocation protocols
- Enable scalable coordination in shared workspaces
- Implement real-world applications in manufacturing assembly
Methodology
Our approach combines distributed planning algorithms with game-theoretic task allocation mechanisms. We use probabilistic roadmaps for spatial planning and develop auction-based protocols for task negotiation. The system employs machine learning to adapt coordination strategies based on team performance and environmental constraints.
Results
Successfully demonstrated coordination of up to 8 robots in shared assembly tasks with 90% efficiency. Reduced task completion time by 45% compared to sequential approaches. The system handles dynamic obstacles and changing priorities with minimal coordination overhead, maintaining scalability as team size increases.
Impact
This research enables more efficient and flexible manufacturing systems where multiple robots can collaborate safely and effectively. Applications include automated assembly lines, warehouse automation, and search and rescue operations requiring coordinated robot teams.
Funding
- Industrial Technology Development Grant — BND $95,000 (2022-2025)
Collaborators
- Dr. Mac Schwager (Stanford University)
- Dr. Vijay Kumar (University of Pennsylvania)
Publications
- Dynamic Task Allocation for Multi-Robot Assembly — A.A. Bhat, M. Schwager, V. Kumar (2023)
- Spatial Planning in Shared Robot Workspaces — A.A. Bhat, V. Kumar (2024)