EMG Versus Torque Control of Human-Machine Systems: Equalizing Control Signal Variability Does not Equalize Error or...

Area of Science:

• Human-computer interaction
• Motor control
• Robotics

Background:

• Understanding the factors influencing human control signals is crucial for designing intuitive and effective human-machine interfaces.
• Electromyography (EMG) and torque are common control signals, but their performance characteristics differ.
• The impact of signal variability on movement error and uncertainty requires further investigation.

Purpose of the Study:

• To investigate whether artificially increasing the variability of torque control signals to match electromyography (EMG) signals results in similar movement errors and uncertainty.
• To compare the performance and confidence associated with EMG and torque control signals.

Main Methods:

• Two experiments were conducted using three control signals: torque, torque with added noise, and EMG.
• Movement error was measured using a target-hitting task with terminal visual feedback.
• Movement uncertainty was quantified by measuring the just-noticeable difference of a visual perturbation.

Main Results:

• Electromyography (EMG) control signals resulted in larger movement errors than torque control signals, even when the signal-to-noise ratio was equalized.
• For equivalent movement errors, EMG control led to higher movement uncertainty compared to both torque and noisy torque control signals.
• Performance and confidence appear to be influenced by factors beyond mere control signal noisiness.

Conclusions:

• Factors such as feedback integration and internal model development significantly impact user performance and confidence.
• Users may struggle to differentiate between random and systematic errors when using EMG control.
• Future research should explore the specific error types associated with EMG control in greater detail.