[8a-N304-1~11] FS.1 Focused Session "AI Electronics"

11:45 AM - 12:00 PM JST(2:45 AM - 3:00 AM UTC)

[8a-N304-11] Research on Skin-Type Tactile Photo Sensing Sheet

〇(M2)Haruto Koizumi¹, Yanchen Shen¹, Satoshi Sunada¹ (1.Kanazawa Univ.)

Keywords:

tactile sensing,machine learning,Skin-Type

This study proposes a skin-like optical tactile sensing method using optical fiber bundles, achieving both a thin profile and wide-area measurement. By externally separating the light-emitting and receiving components, the system estimates the indentation position and depth based on the interruption of hollow optical paths. The adoption of a low-coherence LED prevents interference, while the hollow structure suppresses noise caused by temperature changes, allowing for expected stable measurement. Furthermore, applying linear regression to data processing significantly reduces computational costs while achieving highly accurate estimation results.