Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Automatic Calibration for an Open-source Magnetic Tactile Sensor

Lowiek Van den Stockt, Remko Proesmans, Francis wyffels

TL;DR

The paper presents an open-source magnetic tactile fingertip with a 2×2 taxel grid and an automatic, in situ, gripper-agnostic calibration pipeline that enables immediate use after calibration. The method uses a robot-mounted probe and a force/torque sensor to collect co-registered Hall and force data, then learns per-taxel mappings from polynomial Hall features to the 3D force components $(F_x,F_y,F_z)$ via least-squares regression with an 80/20 train-test split. Hardware design features a resin-dome magnetic assembly and a 2×2 Hall-sensor PCB, while the calibration workflow is designed to be robust to varying robot configurations and readout rates. Results indicate an average $R^2$ of approximately $0.83$ and an average $MSE$ of about $0.97\ N^2$, with further work planned to scale the grid, compensate for hysteresis, and add a second calibration stage for improving coupling effects, thereby lowering barriers to adoption of tactile sensing in robotics.

Abstract

Tactile sensing can enable robots to perform complex, contact-rich tasks. Magnetic sensors offer accurate three-axis force measurements while using affordable materials. Calibrating such a sensor involves either manual data collection, or automated procedures with precise mounting of the sensor relative to an actuator. We present an open-source magnetic tactile sensor with an automatic, in situ, gripper-agnostic calibration method, after which the sensor is immediately ready for use. Our goal is to lower the barrier to entry for tactile sensing, fostering collaboration in robotics. Design files and readout code can be found at https://github.com/LowiekVDS/Open-source-Magnetic-Tactile-Sensor}{https://github.com/LowiekVDS/Open-source-Magnetic-Tactile-Sensor.

Automatic Calibration for an Open-source Magnetic Tactile Sensor

TL;DR

The paper presents an open-source magnetic tactile fingertip with a 2×2 taxel grid and an automatic, in situ, gripper-agnostic calibration pipeline that enables immediate use after calibration. The method uses a robot-mounted probe and a force/torque sensor to collect co-registered Hall and force data, then learns per-taxel mappings from polynomial Hall features to the 3D force components via least-squares regression with an 80/20 train-test split. Hardware design features a resin-dome magnetic assembly and a 2×2 Hall-sensor PCB, while the calibration workflow is designed to be robust to varying robot configurations and readout rates. Results indicate an average of approximately and an average of about , with further work planned to scale the grid, compensate for hysteresis, and add a second calibration stage for improving coupling effects, thereby lowering barriers to adoption of tactile sensing in robotics.

Abstract

Tactile sensing can enable robots to perform complex, contact-rich tasks. Magnetic sensors offer accurate three-axis force measurements while using affordable materials. Calibrating such a sensor involves either manual data collection, or automated procedures with precise mounting of the sensor relative to an actuator. We present an open-source magnetic tactile sensor with an automatic, in situ, gripper-agnostic calibration method, after which the sensor is immediately ready for use. Our goal is to lower the barrier to entry for tactile sensing, fostering collaboration in robotics. Design files and readout code can be found at https://github.com/LowiekVDS/Open-source-Magnetic-Tactile-Sensor}{https://github.com/LowiekVDS/Open-source-Magnetic-Tactile-Sensor.
Paper Structure (7 sections, 4 figures, 1 table)

This paper contains 7 sections, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Sensor design
  3. Automatic Calibration
  4. Hardware Setup
  5. Calibration Methodology
  6. Results
  7. Conclusion

Figures (4)

  • Figure 1: Fingertip structure. The magnets are glued into the domes, the domes are glued onto the PCB. The PCB is pressed between the cover and backing, and the cover is bolted to the backing.
  • Figure 2: Mould for silicone pour. The domes on the mock PCB are shorter than the resin-printed domes, such that after removal of the mock PCB and insertion of the PCB with resin domes, an air cavity is present all around each dome.
  • Figure 3: Calibration probe assembly. The probe tip perfectly fits a taxel dome.
  • Figure 4: F/T Readings and model predictions for taxel 3.