Intention Prediction for Active Upper-Limb Exoskeletons in Industrial Applications: A Systematic Literature Review

Area of Science:

• Robotics and Human-Computer Interaction
• Biomechanics and Wearable Technology

Background:

• Intuitive control of upper-limb exoskeletons is crucial for industrial settings.
• Current intention prediction methods for exoskeletons lack clarity regarding real-world applicability.

Purpose of the Study:

• To systematically review intention prediction in active exoskeletons.
• To identify suitable cues, sensors, and models for industrial applications.
• To provide recommendations for robust and adaptable intention prediction strategies.

Main Methods:

• Systematic literature review following PRISMA guidelines.
• Analysis of 29 studies published between 2007 and 2024.
• Evaluation of sensor modalities (motion capture, electromyography) and prediction approaches (regression, classification).

Main Results:

• Most studies use motion capture and electromyography for prediction.
• Predictions range from 450 ms before to 660 ms after motion onset.
• Limited evaluation of usability and effectiveness in real-world conditions.

Conclusions:

• Recommendations for sensor selection and control strategies are proposed.
• Future research should focus on wearable sensors, cognitive cues, and advanced machine learning.
• Prioritizing real-world validation and diverse participant groups is essential for exoskeleton deployment.