서강대학교 전자공학과

 A paper from AIRLAB accepted to IEEE ICRA’25

From left, Gi Don Han, Jeongwoo Park, and Advisor Prof. Changjoo Nam

A research paper from AI Robotics Lab (directed by Prof. Changjoo Nam) from the Department of Electronic Engineering has been accepted to the International Conference on Robotics and Automation (ICRA 2025), one of the top conferences in robotics and automation. Organized by IEEE, ICRA is the most prestigious international conference in the robotics field and is scheduled to take place from May 19 to 23 at the Georgia World Congress Center in the United States.

The research was led by Ki Don Han, a master's graduate (currently at Samsung Electronics), with contributions from Jeongwoo Park, a Ph.D. candidate. The paper, titled "Stop-N-Go: Search-based Conflict Resolution for Motion Planning of Multiple Robotic Manipulators," presents a novel approach to motion planning in environments with multiple robotic manipulators.

In industrial settings such as factories and warehouses, multiple robotic manipulators often work together on processes such as assembly, welding, painting, and packaging. However, as these robots move simultaneously, they risk colliding with each other. One common approach to solving this problem is to plan all robots' movements in a unified configuration space. While this method considers all robots' movements together, it becomes computationally intractable as the number of robots increases, often failing to find feasible solutions. On the other hand, decoupled planning, where each robot's path is planned independently, is computationally efficient but frequently leads to trajectory conflicts.

The search algorithm for conflict resolution and the environments for experiments

To address this issue, Professor Nam's team proposed a new method that inserts pauses into individually planned trajectories to resolve conflicts. Using the A* algorithm, their approach minimizes the number of stops while reducing the total task completion time. By allowing some robots to stop temporarily, others can move without collisions, ensuring smooth and efficient motion planning. This technique is expected to significantly enhance the efficiency of multi-robot systems operating in shared spaces. Moreover, its applications extend beyond industrial settings to logistics, healthcare, and service robotics, where collaborative robots are increasingly in demand.