Amay Bandodkar, a fourth year PhD student at UCSD, explains that the sensors are programmed to react to the amount of lactate the body produces.
This is an interesting idea. In endurance sports like running, cycling, cross country skiing, etc. there is a parameter that athletes tend to base their training around called "lactate threshold". It's basically the point at which your muscles being producing lactate faster than it can be "buffered" and it is believed this is what causes fatigue and "the burn" you get when you run too fast. You'll notice that if you run or cycle at an endurance pace (below your lactate threshold) you could continue that same effort for hours. When you go above your lactate threshold, you can only sustain that effort for minutes.
Currently the way it is commonly tested is like this (source http://en.wikipedia.org/wiki/Lactate_threshold): "Accurately measuring the lactate threshold involves taking blood samples (normally a pinprick to the finger, earlobe or thumb) during a ramp test where the exercise intensity is progressively increased. Measuring the threshold can also be performed non-invasively using gas-exchange (Respiratory quotient) methods, which requires a metabolic cart to measure air inspired and expired."
In other words you have to measure it in a lab, and you can't really measure it while "on the move". So you can imagine, it might be useful for athletes to just strap a patch onto their skin and have some wireless computer read out the data in real time (like a heart rate monitor)