Decoding Intent With Control Theory: Comparing Muscle Versus Manual Interface Performance

Area of Science:

• Human-Computer Interaction
• Rehabilitation Engineering
• Assistive Technology

Background:

• Manual device interaction demands fine motor skills, posing challenges for individuals with motor impairments.
• Muscle interfaces present an alternative interaction method, but their efficacy in continuous tasks remains unevaluated.
• Control theory offers a framework to analyze user performance by distinguishing intent from error correction.

Purpose of the Study:

• To compare the effectiveness of manual versus muscle interfaces for continuous tasks.
• To quantify the impact of motor impairments on device interaction using control theory.
• To assess user preference and performance across different task complexities.

Main Methods:

• Evaluated manual and muscle interfaces with eleven unimpaired users and three users with motor impairments.
• Utilized control theory to separate user input into intent and error correction components.
• Assessed performance in both simple (e.g., mouse tracking) and complex (e.g., driving) continuous tasks.

Main Results:

• Both unimpaired and motor-impaired users preferred and performed better with the muscle interface.
• The performance advantage of the muscle interface was particularly evident in the complex continuous task (driving).
• Muscle interfaces demonstrated enhanced usability for continuous device control.

Conclusions:

• Muscle interfaces show significant potential for improving device interaction, especially for continuous tasks.
• Algorithms that detect and augment user intent within muscle interfaces could be crucial for future designs.
• These findings suggest muscle interfaces are a promising avenue for enhancing usability for individuals with motor impairments.