A Human-Following Motion Planning and Control Scheme for Collaborative Robots Based on Human Motion Prediction

Area of Science:

• Robotics
• Human-Robot Interaction
• Automation

Background:

• Collaborative robots enhance worker efficiency in assembly lines.
• Ensuring worker safety and comfort during human-robot interaction is crucial.
• Existing systems often lack dynamic adaptation to worker movement.

Purpose of the Study:

• To develop a human-following motion planning and control scheme for collaborative robots.
• To enable robots to dynamically supply parts and tools to moving workers.
• To ensure worker safety and comfort in an industrial assembly setting.

Main Methods:

• Utilized a 3-D sensing system to capture real-time worker skeletal data.
• Employed probabilistic prediction of future worker positions.
• Implemented a Model Predictive Control (MPC)-based trajectory planner for robot motion.
• Integrated the scheme into a 2-DOF planar manipulator collaborative robot system.

Main Results:

• The proposed scheme successfully enabled the collaborative robot to follow a moving worker.
• The robot effectively supplied necessary parts and tools to the worker.
• Experimental results demonstrated enhanced safety and comfort for the worker.
• The system provided continuous assistance to workers moving within the workspace.

Conclusions:

• The developed human-following system enhances collaborative robot functionality in dynamic assembly tasks.
• The integration of real-time sensing, predictive control, and MPC ensures safe and comfortable human-robot collaboration.
• This approach offers a significant advancement in adaptive robotic assistance for industrial environments.