Abdullah Alothman
David Boe
Niveditha Kalavakonda

Recent efforts to integrate sensory feedback into upper limb prostheses have focused on substituting exteroceptive sensations like touch and texture. However, little has been done to address the idea that feedback and control of a prosthesis are exquisitely linked. In fact, the nature of using EMG signals as a controller introduces uncertainty into the system. We are testing the hypothesis that substituting interoceptive sensation enhances one’s ability to manipulate an uncertain control signal.

We test our subjects on a task that simulates the real world control demands of a myoelectric prosthesis, in which the subject is asked to hit targets using forearm muscle activation. We can detect muscle activity with off-the-shelf sensors and convert the resultant EMG signal into feedback using an Arduino and vibrotactile actuators mounted on an arm band. The subject can use this feedback to determine what level of activation they have achieved, and thus how close to the target level they are. We have integrated our hardware with MATLAB and Unity to continue exploring this domain!

We are currently collecting and analyzing data from human subjects, and we look forward to providing an update soon.


Mentored by:

Eric Rombokas