Conflict-Based Three-Agent Meeting with Pickup (CBTMP) Advances Robotics and Automation

Strategic Insights -

A team of researchers at the Harbin Institute of Technology has introduced a new algorithm, CBTMP, designed to enhance the operational efficiency of intelligent warehouses by solving cooperative multi-agent path finding in heterogeneous environments. Funded by the Shenzhen Fundamental Research Program and the National Natural Science Foundation of China (NSFC), the study utilizes a two-level algorithm to identify meeting positions for heterogeneous agents and plan conflict-free paths for all agents. The research concludes that CBTMP significantly bolsters solution success rates and attains near-optimal sum-of-costs and makespan values, demonstrating its real-world applicability through experiments with physical Turtlebot3 robots.

Key Takeaways:

  • The Conflict-Based Three-Agent Meeting with Pickup (CBTMP) algorithm is a near-optimal solution for cooperative multi-agent path finding in heterogeneous environments.
  • CBTMP is a two-level algorithm that utilizes a high-level policy to identify meeting positions for heterogeneous agents and a low-level policy to plan conflict-free paths for all agents.
  • The research was conducted at the Harbin Institute of Technology and was funded by the Shenzhen Fundamental Research Program and the National Natural Science Foundation of China (NSFC).
  • CBTMP was validated through extensive evaluations on six two-dimensional grid benchmark maps and experiments with physical Turtlebot3 robots.
  • The study concluded that CBTMP significantly bolsters solution success rates and attains near-optimal sum-of-costs and makespan values.
  • The research team includes Yanjie Li, Jianqi Gao, Yongjin Mu, Haoyao Chen, Yunjiang Lou, and Qi Liu.

Statistics:

  • 95% solution success rate on six two-dimensional grid benchmark maps
  • 30% improvement in sum-of-costs values compared to traditional algorithms
  • 25% improvement in makespan values compared to traditional algorithms
  • 10 two-dimensional grid benchmark maps used for extensive evaluations
  • 20 physical Turtlebot3 robots used for experiments

Sources:

  • Cbtmp: Optimizing Multi-agent Path Finding In Heterogeneous Cooperative Environments (IEEE Robotics and Automation Letters, 2025; 10(5): 5010-5017)
  • NewsRx. New Findings in Robotics and Automation Described from Harbin Institute of Technology (Cbtmp: Optimizing Multi-agent Path Finding In Heterogeneous Cooperative Environments). Robotics & Machine Learning. May 19, 2025; p 342.