Sensor-based robot deployment algorithms

In robot deployment problems, the fundamental issue is to optimize a steady state performance measure that depends on the spatial configuration of a group of robots. For such problems, a classical way of designing high-level feedback motion planners is to implement a gradient descent scheme on a suitably chosen objective function. This can lead to computationally expensive algorithms that may not be adaptive to uncertain dynamic environments. We address these challenges by showing that algorithms for a variety of deployment scenarios in uncertain stochastic environments and with noisy sensor measurements can be designed as stochastic gradient descent algorithms, and their convergence properties analyzed via the theory of stochastic approximations. This approach yields often surprisingly simple algorithms that can accommodate complicated objective functions, and work without a detailed model of the environment. To illustrate the richness of the framework, we discuss two applications, namely source seeking with realistic stochastic wireless connectivity constraints, and coverage with heterogeneous sensors.