Demo: Coordinated Actuation for Sensing Uncertainty Reduction

We are interested in improving the sensing fidelity achieved using a given set of sensing resources, by aligning the network configuration precisely to the sensing requirements in the deployment environment. For example, consider a security application where cameras are deployed to monitor all humans present at a shopping mall. A naïve strategy would be to install enough sensors to cover the entire volume of the shopping mall at the resolution required for recognizing faces. However, if we consider the properties of the phenomenon being sensed, and realize that only a very small fraction of the volume actually consists of regions of interest (in this case, faces) we may be able to obtain the same sensing performance with significantly reduced sensor installation, maintenance and data processing costs.