It’s the first and only software that is capable of simulating highly realistic, personalized datasets of electromyography (EMG) signals of arbitrary size and variability. Watch these short videos to see its potential.

Decoding and interpreting EMG signals requires building AI-based algorithms, which need to be trained on a large amount of high quality labelled EMG data. The only way to do it today – using real EMG measurements.

Acquiring real EMG data is expensive, the data quality is low and, most importantly, it usually can not be perfectly labelled.

Using our Myoelectric Digital Twin, we are able to generate arbitrary large datasets of synthetic EMG signals in a very fast and inexpensive way. The data has high variability and quality, and it can be perfectly labelled !

This data can be then used for training AI-based muscular interfaces