Experimental Validation of Modified BP-RRT* for Dual-Arm Collaborative Robots in ROS: A Step Towards Real-Time Deployment

Abstract:
Path planning for dual-arm collaborative robots poses significant challenges due to highdimensional configuration spaces and real-time execution constraints. In our previous research,
we proposed the Modified BP-RRT* algorithm, demonstrating its superiority in generating
quicker and more efficient collision-free trajectories. This study focuses on the experimental
validation of the algorithm within the Robot Operating System (ROS) framework to assess its
practical feasibility in real-world robotic applications. The algorithm is integrated with a dualarm robotic system in a simulated ROS environment, and its performance is evaluated against
conventional sampling-based planners. Key metrics such as path smoothness, computational
efficiency, execution time, and success rate are analysed. The results confirm that the Modified
BP-RRT* significantly reduces planning time while maintaining optimality and collision
avoidance, making it well-suited for dynamic and industrial environments. This validation
bridges the gap between theoretical advancements and real-world robotic deployment, paving
the way for future applications in robotic assembly, human-robot collaboration, and
autonomous manipulation