Systems, Algorithms and Applications for Mobile Robots in Environmental Sustainability

Mobile robots have become ubiquitous today. Advances in on-board computation, hardware design and sensor accuracy have propelled the design of autonomous mobile robots for deployment in real-world applications. Forests, agricultural farms, pastures and grasslands are the largest land use categories in the US and can benefit immensely from the use of autonomous mobile robots. The scale and complexity of operating in these uncontrolled or semi-controlled settings create a need to design algorithms to build and improve autonomous robot capabilities. In this talk, I will present two applications for mobile robot systems to aid in environmental monitoring and to provide sustainable cost-effective agricultural solutions. The first application involves large scale area coverage using aerial imagery. While aerial robots are used extensively for mapping, monitoring and surveillance, their range of operation constraints their use. I will present cooperative planning algorithms to use a ground-based mobile refueling robot to increase the spatio-temporal range of a fuel-constrained unmanned aerial vehicle (UAV). The second application addresses weed removal in organic dairy pastures using an autonomous mobile robot. The carbon footprint of using diesel-run farm vehicles for weed removal and other agricultural tasks has been a cause of concern, especially in the case of organic farms that do not use chemicals. I will present the design of Cowbot (autonomous weed mowing robot for cow pastures) and online planning algorithms to compute efficient trajectories for the Cowbot that utilize detection information from its onboard sensing systems.