Bio-inspired Multi-Agent RL (MARL)

Tasks:

• Develop and train DL models to simulate large “active” clusters (>100 agents)

• Design control strategies for practical applications

• Reveal and analyze emergent collective dynamics in robot swarms

• Explore decentralized RL for emergent coordination 

Solving practical problems for:

• Coordinated cargo transport

• Traffic flow optimization (warehouse scenarios)

• Decentralized obstacle avoidance and navigation