Human-Robot Interaction Using Learning from Demonstrations and a Wearable Glove with Multiple Sensors

Area of Science:

• Robotics
• Human-Computer Interaction
• Machine Learning

Background:

• Effective human-robot interaction (HRI) is crucial for seamless collaboration and enhanced productivity.
• Complex tasks like robotic grasping require integrating human intelligence with robotic capabilities.
• Current HRI methods need improvement for nuanced tasks.

Purpose of the Study:

• To enhance human-robot interaction for complex tasks.
• To develop a system for collecting human motion and force data using a wearable glove.
• To enable robots to understand and replicate human manipulation skills.

Main Methods:

• A wearable glove equipped with accelerometer, pressure, and flex sensors was utilized.
• Multi-sensor fusion techniques were employed to gather comprehensive human motion and force data.
• A machine learning algorithm was developed for analyzing sensor data.

Main Results:

• The system successfully collected motion and force data relevant to object handling.
• The machine learning algorithm demonstrated the ability to recognize grasp orientation and position.
• Sensor fusion provided rich data for nuanced robotic manipulation tasks.

Conclusions:

• Wearable sensor technology can significantly improve human-robot interaction.
• Machine learning analysis of multi-sensor data enables robots to interpret human intent for grasping.
• This approach facilitates more intuitive and effective human-robot collaboration in complex manipulation tasks.