Intrinsic sense of touch for intuitive physical human-robot interaction.

Area of Science:

• Robotics
• Human-Computer Interaction
• Artificial Intelligence

Background:

• The sense of touch is crucial for delicate environmental interaction.
• Current human-robot interaction often lacks intuitive tactile feedback.
• Robotic systems typically rely on external sensors or artificial skin for touch.

Purpose of the Study:

• To develop an intrinsic robotic sense of touch.
• To enable intuitive human-robot interaction without artificial instrumentation.
• To interpret human touch as machine-readable commands.

Main Methods:

• Utilizing high-resolution joint-force-torque sensing in a redundant configuration.
• Applying manifold learning techniques and artificial neural networks.
• Developing a concept for 'virtual buttons' for tactile communication.

Main Results:

• Robots can sensitively perceive and localize human touch trajectories.
• Touch trajectories are interpreted as letters, symbols, or numbers.
• Virtual buttons, sliders, and switches were implemented for tactile communication.

Conclusions:

• An intrinsic sense of touch advances physical human-robot interaction.
• This technology offers intuitive, flexible, and adaptable interaction modalities.
• It moves beyond conventional interaction methods towards more natural handling.