Synthetic Tactile Sensor for Macroscopic Roughness Estimation Based on Spatial-Coding Contact Processing.

Area of Science:

• Robotics and Artificial Intelligence
• Biomimetics and Haptics
• Materials Science and Engineering

Background:

• Traditional tactile sensors lack the ability to quantify human perception of surface roughness.
• Human tactile perception relies on processing subcutaneous deformation and neural signals from textured surfaces.
• Bridging the gap requires sensors that emulate human tactile information processing.

Purpose of the Study:

• To develop a method for predicting macroscopic roughness perception using a human-like tactile sensing principle.
• To replicate the spatial-coding mechanism of human tactile assessment.
• To validate a vision-based approach for estimating human roughness perception.

Main Methods:

• A camera captured contact information through a flexible, transparent material with fingerprint-like structures.
• Contact images were recorded under varying forces (1 N to 3 N).
• Spatial frequency components (0.1-1.0 mm⁻¹) were extracted and linearly combined to approximate human perception.

Main Results:

• The developed method successfully predicted macroscopic roughness for surfaces with protrusions (wavelengths 2-5 mm).
• Estimated roughness values showed an average error comparable to human perception variability.
• The approach demonstrated effective replication of human roughness perception.

Conclusions:

• Vision-based tactile sensing can effectively replicate human macroscopic roughness perception.
• The proposed method offers a promising approach for developing advanced tactile sensors.
• This research advances biomimetic sensing and human-computer interaction.