# Robust Affordable 3D Haptic Sensation via Learning Deformation Patterns

## TL;DR

This paper introduces a machine learning-based framework for inferring 3D surface forces from internal deformations using minimal sensors, achieving high accuracy and optimized sensor placement.

## Contribution

It presents a novel data-driven approach for sparse sensor placement and force inference on 3D structures, reducing sensor count while maintaining precision.

## Key findings

- Achieved 8 mm localization precision on a robotic limb.
- Optimized sensor placement using finite element simulation data.
- Compared data-driven methods with geometric and information-based criteria.

## Abstract

Haptic sensation is an important modality for interacting with the real world. This paper proposes a general framework of inferring haptic forces on the surface of a 3D structure from internal deformations using a small number of physical sensors instead of employing dense sensor arrays. Using machine learning techniques, we optimize the sensor number and their placement and are able to obtain high-precision force inference for a robotic limb using as few as 9 sensors. For the optimal and sparse placement of the measurement units (strain gauges), we employ data-driven methods based on data obtained by finite element simulation. We compare data-driven approaches with model-based methods relying on geometric distance and information criteria such as Entropy and Mutual Information. We validate our approach on a modified limb of the Poppy robot and obtain 8 mm localization precision.