How can AI reduce wrist injuries in the workplace?

2025-05-30

AI Summary

To address the high incidence of occupational wrist injuries in industrial settings, this study proposes a lightweight, adaptive wrist exoskeleton system. Methodologically, we introduce a low-complexity sensing scheme—comprising only an 8-channel surface electromyography (sEMG) array and a handheld force gauge—and develop a joint control model for simultaneous gesture classification and force estimation. The model integrates sEMG signal acquisition, pattern recognition, and regression modeling, culminating in a fully integrated industrial wearable prototype. Evaluation on data from six production-line workers achieves a gesture classification accuracy of 94.2% and a mean absolute force prediction error of <1.8 N. Compared to existing solutions, the prototype features simplified mechanical design, sub-80-ms control latency, and over 30% cost reduction, thereby significantly enhancing system reliability and feasibility for large-scale industrial deployment.

Technology Category

• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Computer Vision: Biometrics, Face, Gesture & Pose

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

This paper explores the development of a control and sensor strategy for an industrial wearable wrist exoskeleton by classifying and predicting workers' actions. The study evaluates the correlation between exerted force and effort intensity, along with sensor strategy optimization, for designing purposes. Using data from six healthy subjects in a manufacturing plant, this paper presents EMG-based models for wrist motion classification and force prediction. Wrist motion recognition is achieved through a pattern recognition algorithm developed with surface EMG data from an 8-channel EMG sensor (Myo Armband); while a force regression model uses wrist and hand force measurements from a commercial handheld dynamometer (Vernier GoDirect Hand Dynamometer). This control strategy forms the foundation for a streamlined exoskeleton architecture designed for industrial applications, focusing on simplicity, reduced costs, and minimal sensor use while ensuring reliable and effective assistance.

Problem

• Develop AI-driven wrist exoskeleton to prevent workplace injuries
• Classify and predict workers' actions using EMG-based models
• Optimize sensor strategy for cost-effective industrial exoskeleton design

Innovation

• EMG-based wrist motion classification model
• Force regression model using dynamometer data
• Simplified exoskeleton with minimal sensors