Hi everybody,

I am interested in examples of phenomena (values measured by sensors)
that can be predicted somehow. For example, in the context of
vehicles, the location of the vehicle can be predicted by taking into
account features of the road such as the average speed and the route.
Thus, for example, I can approximate the X-location of a vehicle with
a function such as loc_x(t) = loc_x(0) + v * t, assuming a straight
road. This function allows me to do computations about locations at
any time instant (e.g., get the number of vehicles within a certain
area) as long as the real value does not change too much.

My goal is to find application examples where the above idea can be
applied: using predictions to save computational effort in getting
answers. I assume that when the real situation changes too much, a new
function will be considered to estimate future values.

The context of moving objects (where the sensors are GPS receivers) is
the one I am involved, but I guess there must be many more interesting
similar applications... and I need to find some of them. Your help is
greatly welcome!