Elastomer-Based Visuotactile Sensor for Normality of Robotic Manufacturing Systems

Area of Science:

• Robotics
• Materials Science
• Manufacturing Engineering

Background:

• Achieving high accuracy and reliability in aircraft assembly is crucial, with drilled hole normality being a critical geometrical tolerance.
• Current elastomer-based tactile sensors lack the precision required for robotic machining in aerospace applications.
• Failure to meet tight tolerances can lead to structural fatigue and assembly errors.

Purpose of the Study:

• To develop a novel elastomer-based tactile sensor for precision cobot machining in aerospace.
• To characterize elastomer materials for optimal deformability in tactile sensing.
• To validate the sensor's performance in achieving critical geometrical tolerances.

Main Methods:

• Mechanical testing of three commercial silicon-based elastomer materials.
• Finite element modeling to simulate sensor deformation on surfaces with varying normalities.
• Additive manufacturing for mould fabrication, chemical etching for surface improvement, and direct casting for sensor production.
• Machine learning for training simulated and experimental sensor data.

Main Results:

• Selection of an optimal elastomer material based on deformability.
• Successful fabrication of the tactile sensor using additive manufacturing and direct casting.
• A mean perpendicularity tolerance of 0.34° was achieved in real-world experiments.
• The developed sensor demonstrated capability in evaluating various inclination angles.

Conclusions:

• The novel elastomer-based tactile sensor is suitable for precision cobot machining.
• The developed sensor offers a new perspective for low-cost, high-accuracy robotic assembly in the aerospace industry.
• This technology addresses the limitations of current tactile sensors in achieving tight geometrical tolerances.