Master's Thesis

Design of a Highly Dense Tactile Fingertip for Dexterous Hands

Miguel de Oliveira Lameiras — 2025

Key information

Authors:

Miguel de Oliveira Lameiras (Miguel de Oliveira Lameiras)

Supervisors:

Susana Isabel Pinheiro Cardoso de Freitas (Susana Isabel Pinheiro Cardoso de Freitas); Plinio Moreno López (Plinio Moreno Lopez)

Published in

November 17, 2025

Abstract

Tactile sensing for dexterous robotic manipulation requires sensors that simultaneously achieve high spatial resolution, sensitivity, and compact form factor, capabilities that remain difficult to combine in existing solutions. This thesis presents the design, fabrication, and experimental validation of a highdensity tactile fingertip, integrating seven Tunneling Magnetoresistance (TMR) sensors beneath a Magnetorheological Elastomer (MRE) with a 4.86 cm2 contact surface. The fingertip was developed as a compact and scalable tactile sensing platform with geometry and dimensions closely matching those of a human fingertip, capable of estimating both contact force and contact location. TMR sensor dies, microfabricated at INESC-MN, were directly wire-bonded onto a flexible PCB to achieve higher spatial density than packaged commercial sensors. Each fingertip incorporates embedded signal processing via an STM32G4 microcontroller, with onboard digital filtering, 500 Hz output sample rate, and I2C communication for multi-fingertip scalability. Magnetic annealing of the MRE proved critical, improving magnetic uniformity and yielding a maximum sensitivity of 35.67 mV/N with 72 dB SNR. A calibration framework was developed to map sensor responses to contact parameters without requiring robotic positioning systems. A neural network trained on approximately 80000 contact samples achieved mean absolute errors of 0.32 N for force and 0.023/0.016 mm for x-y localization at training positions. When evaluated on 12 unseen contact locations, localization remained within 1.0 mm, comparable to human two-point discrimination, while force estimation maintained errors below 0.65 N, demonstrating effective spatial generalization.

Publication details

Authors in the community:

Supervisors of this institution:

Fields of Science and Technology (FOS)

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

Publication language (ISO code)

eng - English

Rights type:

Embargoed access

Date available:

November 13, 2026

Institution name

Instituto Superior Técnico