The quintessential example is probably detecting the presence of an object in front of a light sensor in a way that works regardless of the ambient light in the room. There are lots of other examples, though, e.g. a hand-made pressure or stretch sensor may not always return to the same baseline after activation. Capacitive sensing is another one, as the baseline readings can easily change if the electrode is moved slightly.
There are, I think, lots of potential approaches to doing this: looking for a slope (over some time interval), maintaining an updating baseline, looking for a threshold of change relative to the standard deviation over some previous interval, etc.
The quintessential example is probably detecting the presence of an object in front of a light sensor in a way that works regardless of the ambient light in the room. There are lots of other examples, though, e.g. a hand-made pressure or stretch sensor may not always return to the same baseline after activation. Capacitive sensing is another one, as the baseline readings can easily change if the electrode is moved slightly.
There are, I think, lots of potential approaches to doing this: looking for a slope (over some time interval), maintaining an updating baseline, looking for a threshold of change relative to the standard deviation over some previous interval, etc.