Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A modular architecture for IMU-based data gloves

Alessandro Carfì, Mohamad Alameh, Valerio Belcamino, Fulvio Mastrogiovanni

TL;DR

The paper addresses the reproducibility and standardization gap in IMU-based data gloves for hand tracking. It proposes a modular open hardware/software architecture with core and sensory modules connected via I2C and Flexible Flat Cable interfaces, enabling scalable sensor counts and easy repair. The implementation uses ESP32-based cores and MPU-9250 IMUs with an I2C multiplexer, plus software for sensor data collection and PC streaming, demonstrating eleven sensors operating at over 20 Hz for about an hour. Results indicate that orientation drift is the dominant error source but can be mitigated with software, supporting a practical, extensible platform for future hand-tracking research.

Abstract

The flexibility and range of motion in human hands play a crucial role in human interaction with the environment and have been studied across different fields. Researchers explored various technological solutions for gathering information from the hands. These solutions include tracking hand motion through cameras or wearable sensors and using wearable sensors to measure the position and pressure of contact points. Data gloves can collect both types of information by utilizing inertial measurement units, flex sensors, magnetic trackers for motion tracking, and force resistors or touch sensors for contact measurement. Although there are commercially available data gloves, researchers often create custom data gloves to achieve the desired flexibility and control over the hardware. However, the existing literature lacks standardization and the reuse of previously designed data gloves. As a result, many gloves with unclear characteristics exist, which makes replication challenging and negatively impacts the reproducibility of studies. This work proposes a modular, open hardware and software architecture for creating customized data gloves based on IMU technology. We also provide an architecture implementation along with an experimental protocol to evaluate device performance.

A modular architecture for IMU-based data gloves

TL;DR

The paper addresses the reproducibility and standardization gap in IMU-based data gloves for hand tracking. It proposes a modular open hardware/software architecture with core and sensory modules connected via I2C and Flexible Flat Cable interfaces, enabling scalable sensor counts and easy repair. The implementation uses ESP32-based cores and MPU-9250 IMUs with an I2C multiplexer, plus software for sensor data collection and PC streaming, demonstrating eleven sensors operating at over 20 Hz for about an hour. Results indicate that orientation drift is the dominant error source but can be mitigated with software, supporting a practical, extensible platform for future hand-tracking research.

Abstract

The flexibility and range of motion in human hands play a crucial role in human interaction with the environment and have been studied across different fields. Researchers explored various technological solutions for gathering information from the hands. These solutions include tracking hand motion through cameras or wearable sensors and using wearable sensors to measure the position and pressure of contact points. Data gloves can collect both types of information by utilizing inertial measurement units, flex sensors, magnetic trackers for motion tracking, and force resistors or touch sensors for contact measurement. Although there are commercially available data gloves, researchers often create custom data gloves to achieve the desired flexibility and control over the hardware. However, the existing literature lacks standardization and the reuse of previously designed data gloves. As a result, many gloves with unclear characteristics exist, which makes replication challenging and negatively impacts the reproducibility of studies. This work proposes a modular, open hardware and software architecture for creating customized data gloves based on IMU technology. We also provide an architecture implementation along with an experimental protocol to evaluate device performance.
Paper Structure (7 sections, 2 figures)

This paper contains 7 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Hardware Architecture
  3. Implementation
  4. Results
  5. Conclusions
  6. Acknowledgments.
  7. Disclosure of Interests.

Figures (2)

  • Figure 1: The image compares the architecture on the left with the implementation on the right. The core module is in blue, while the sensory module is in red.
  • Figure 2: On the left, we have the dataglove worn by a human, and on the right, it is mounted on the AR10 Robotic hand.