Controlling Games and Apps Through Muscle Sensors 47
A team with members from Microsoft, the University of Toronto, and the University of Washington have developed an interface that uses electrodes to monitor muscle signals and translate those into commands or button presses, allowing a user to bypass a physical input device and even control a game or application while their hands are full. The video demonstration shows somebody playing Guitar Hero by making strumming motions and tapping his fingers together, a jogger changing his music without having to touch the device, and a man flexing a muscle to open the trunk of his car while he carries objects in both hands. The academic paper (PDF) is available online.
Sensitivity (Score:3, Insightful)
This works fine for off/on states, but not graduated ones where a range of input is needed. Muscles are binary -- they are off, or on. At least, at the cellular level. But when they're put in bunches, only some are activated while others are not, which leads to a range of possible force levels. Effectively monitoring neural activity here requires a large number of sensors to accurately determine how much force is being requested and then translate that into a digital representation. As well, do not forget that in the human body, motion is comprised of two separate inputs from the sympathetic and parasympathetic nervous system: And while complementary, these two are not always perfectly in balance. This is why prothetic limbs have to be computer-assisted and lack fine motor control: They simply can't get a good enough input resolution.
So yes, it'll be great for mouse clicks (binary), but I'll still own your ass in a video game in anything that uses a vector (analog).