At the intersection of robotics, machine learning, and environmental sensing lies a focused research effort on teaching autonomous systems how to sample and reconstruct spatial phenomena. Working in the DREAMS Lab at Arizona State University, Bharath Desikan develops algorithms that enable robots to build accurate maps of environmental fields.
The work centers on spatial field reconstruction using probabilistic models that allow autonomous platforms to reason about the structure of an environment from sparse measurements. These methods are designed for practical deployment, with an emphasis on efficiency and real world constraints.