Related Experiment Video
Updated: Jul 16, 2026
11:53
The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy
Published on: October 14, 2017
Robust Multi-Agent Path Finding Method for Obstacles and Environmental Changes in Factory Environments.
Seihoon Park1, Jinwon Lee1, Geonhyeok Park1
1Department of Mechanical Engineering, Korea University, Seoul 02841, Republic of Korea.
Sensors (Basel, Switzerland)
|July 15, 2026
Summary
This study introduces a new multi-robot path planning framework that adapts to real-time obstacles, improving efficiency and reducing collisions in dynamic factory environments. The obstacle-aware selective replanning method enhances robot fleet operation.
Area of Science:
- Robotics and Automation
- Artificial Intelligence
- Operations Research
Background:
- Multi-Agent Path Finding (MAPF) is crucial for logistics automation.
- Existing guidance-based MAPF methods assume static environments, leading to issues with unexpected obstacles.
- Previous robust MAPF methods (e.g., kR-MAPF) use global parameters that can be inefficient.
Purpose of the Study:
- To develop a MAPF framework that enhances robustness against dynamic obstacles in industrial environments.
- To reduce computational overhead and improve path efficiency compared to existing methods.
- To ensure stable large-scale multi-robot operation.
Main Methods:
- Proposed a multi-robot path planning framework with real-time obstacle-aware selective replanning.
- Identified robots affected by obstacles and selectively replanned their paths.
- Updated guidance policies dynamically based on real-time obstacle information.
Main Results:
- Achieved a 100% success rate in simulations with up to 100 robots in a 100m×100m factory environment.
- Reduced planning runtime by 35-79% compared to kR-MAPF.
- Decreased flowtime by 7-24% compared to kR-MAPF.
Conclusions:
- Obstacle-aware selective replanning significantly improves real-time performance and path efficiency in dynamic factory settings.
- The proposed framework offers a technical basis for stable, large-scale multi-robot operations.
- This approach mitigates unnecessary conservatism and search space increase associated with global robustness parameters.