Achieving efficient multi-sensory perception through predictive correction of monitoring errors

We present a novel approach to managing the quality and cost of perception in a multi-sensor robotic system. The approach involves prediction of monitoring errors of a range of process monitors using a weighted least squares algorithm, and minimization of the cost of monitoring constrained by the desired level of monitoring performance. Two significant characteristics of the approach are: (1) dynamic, real-time management of process monitors, and (2) the ability to deliver high quality information while keeping the cost of perception low