Design and Calibration of a Force/Tactile Sensor for Dexterous Manipulation

Area of Science:

• Robotics
• Sensor Technology
• Human-Computer Interaction

Background:

• Robotic grasping requires advanced sensory feedback beyond simple force detection.
• Existing sensors often struggle with detecting torsional moments and managing soft contact, limiting dexterity.
• Fragile object manipulation necessitates precise control to avoid damage.

Purpose of the Study:

• To design and calibrate a new force/tactile sensor for robotic applications.
• To equip robotic grippers with a human-like sense of touch, including force, slip, and geometry perception.
• To address the challenge of sensing torsional moments and enabling soft distributed contact for secure grasping.

Main Methods:

• Development of a novel sensor design integrating force, slip, and geometry sensing capabilities.
• Implementation of calibration procedures specifically addressing the challenges posed by soft distributed contact.
• Experimental validation using two sensors mounted on an industrial gripper during real grasping tasks.

Main Results:

• The designed sensor successfully mimics human tactile perception, providing crucial information for robotic manipulation.
• The sensor effectively detects contact forces, object slip, and object geometry.
• Calibration methods were developed to handle the complexities of soft contact, enhancing sensor reliability.

Conclusions:

• The new tactile sensor significantly enhances robotic grasping capabilities, particularly for fragile objects and complex manipulation tasks.
• The sensor's ability to perceive torsional moments and facilitate soft contact is critical for secure and dexterous object handling.
• The developed calibration techniques provide a foundation for deploying advanced tactile sensing in real-world robotic applications.