Abstract: This talk considers optimal deployment problems for networks of autonomous robotic sensors and examines their connection with spatial estimation. Given a spatial random field over a region of interest, robotic sensors can improve the efficiency of data collection, adapt to changes in the environment, and provide a robust response to individual failures. We illustrate ways in which systems and control can help us design coordination algorithms to cooperatively optimize data collection, minimize the uncertainty of the estimation, account for individual failures in communication, and deal with uncertainty in the state of the network. Our technical approach provides correctness and performance guarantees by combining ideas and tools from geometric optimization, spatial statistics, self-triggered control and nonsmooth analysis.