Geometry Optimisation of a Hall-Effect-Based Soft Fingertip for Estimating Orientation of Thin Rectangular Objects

Area of Science:

• Robotics
• Materials Science
• Sensor Technology

Background:

• Soft tactile sensors are crucial for robotic grippers in assembly tasks.
• Current Hall-effect sensor design relies on trial-and-error, limiting performance.
• Accurate contact information and object orientation are challenging to achieve during grasping.

Purpose of the Study:

• To investigate optimal geometrical designs for cylindrical soft tactile sensors.
• To enhance sensor sensitivity for improved robotic grasping capabilities.
• To reduce errors in object orientation estimation during robotic manipulation.

Main Methods:

• Finite element modeling (FEM) of a soft fingertip in Abaqus.
• Optimization using Isight, maximizing magnet displacement as the objective function.
• Fabrication and experimental validation of optimized and non-optimized sensor designs.

Main Results:

• Optimal sensor design identified at the boundary of the parameter space.
• Experimental validation showed reduced average error in object orientation estimation.
• Successful estimation of 3-axis forces using the optimized sensor outputs.

Conclusions:

• The proposed method effectively optimizes soft tactile sensor geometry for enhanced performance.
• Optimized sensors significantly improve object orientation estimation accuracy in robotic grasping.
• The developed sensors show potential for precise force sensing in assembly applications.