Human Arm Motion Prediction for Collision Avoidance in a Shared Workspace.

Area of Science:

• Robotics
• Artificial Intelligence
• Industrial Automation

Background:

• Industry 4.0 emphasizes human-centric and digitized industrial systems.
• Increased human-robot collaboration necessitates robust safety and efficiency measures.
• Traditional systems lack the adaptability for dynamic, shared workspaces.

Purpose of the Study:

• To develop a framework for robots to perceive, predict, and react to human motion.
• To minimize collisions in shared human-robot workspaces.
• To enhance safety and efficiency in collaborative industrial environments.

Main Methods:

• Equipping robots with exteroceptive sensors for environmental perception.
• Utilizing an Encoder-Decoder network for human hand trajectory prediction.
• Implementing a Model Predictive Control (MPC) framework for collision-free motion planning.

Main Results:

• The proposed system effectively predicts human trajectories in real-world shared workspaces.
• The MPC framework successfully generated collision-free robot paths based on predictions.
• Validated framework ensures safe human-robot collaboration.

Conclusions:

• The integration of predictive human motion tracking and MPC enables safe and efficient human-robot collaboration.
• This approach is crucial for advancing Industry 4.0 applications.
• The system demonstrated practical viability in a real-world setting.