News Facebook acquires BCI specialist CTRL-Labs

I was about to mention this sounds like the old Myo wristband (a total failure, the concept of viable EMG from a slidey-aboutey wristband is a non-starter), but it looks like CTRL-Labs acquired Myo's patents earlier this year so everything old is new again.

The root problem is that useful EMG requires proper electrode placement above the muscle (or more often, muscle group) you are measuring. Misplacement of the electrode (e.g. if it is applied to a band that a user slides on in whatever orientation and position they find comfortable, rather than applied to the skin by a nurse or technical or applied to the inside of a stump cup) means you are not measuring the muscle(s) you think you are measuring. At best, you can give some dedicated training tasks intended to produce known activation patterns (e.g. move you first finger, move your second finger, extend your wrist, flex your wrist, etc) but not only is this tedious and unreliable, it also is invalidated as soon as the band slips or is moved (because it becomes sweaty and uncomfortable) and needs to be retrained. Plus any muscles layered beneath other muscles (e.g. the Abductor Pollicis Longus tucked under the Supinator on one side and the Extensor Digitorum on the other) have their activation signals obscured by those overlaying muscle's activating.