Santosh Pitla, University of Nebraska-Lincoln, USA: Development of Control Architectures for Multi-Robot Agricultural Field Production Systems
The evolution of agricultural equipment from simple mechanical devices to state-of-the-art machines capable of performing tasks autonomously is on the horizon. Modern autonomous machines, although equipped with numerous sensors and controls, require human supervision to warrant safe operation. In contrast, the next generation autonomous machines, in addition to the control hardware, must include intelligence to learn and react to the unknown environment in the absence of an operator. Intelligence to an Agricultural Robot (Ag-Robot) is possible through a robust control methodology/architecture that effectively interprets, arbitrates and prioritizes the information obtained from sensors to produce desired actions.
This talk presents the Individual Robot Control Architecture (IRCA) and Multi Robot System Control Architecture (MRSCA), developed to provide intelligence to an individual and a group of coordinated agricultural robots, respectively. The architectures were simulated using MATLAB tools SIMULINK and STATEFLOW CHART, and were deployed on three Autonomous Vehicle Platforms (AVPs) for validation. The successful testing of the AVP in two test scenarios—autonomous single vehicle navigation and autonomous single vehicle obstacle detection and avoidance—validated the Deliberative and Reactive features of the IRCA. The successful simulation of the planting, baling-bale retrieving and the harvest operations using the AVPs in three test scenarios validated the Standalone, Modest and Absolute level of cooperation offered by the MRSCA.
Dr. Santosh Pitla got the B.S. in Mechanical Eng. from Osmania University, India, in 2004 and the M.Sc. in Biosystems and Agricultural Engineering from the University of Kentucky in 2007. For his master’s project he developed a tractor-mounted instrumented Coulter disc that can oscillate and map compact layers at various depths in the soil. After finishing that work, he became a Research Engineer at University of Kentucky while simultaneously working on a part-time Ph.D. With his Ph.D. dissertation work, he got involved in mobile agricultural robotic research. He developed CAN-based small-scale autonomous vehicle platforms to evaluate the control methodologies that were developed. He graduated in the spring of 2012 and moved to the Ohio State University as Post-doctoral scholar. His Post-Doc work dealt with modeling the economics of autonomous agricultural machines relative to conventional machines for corn and soybean production. Since October 2013 he is an assistant professor in the Department of Biological Systems Engineering at the University of Nebraska-Lincoln. His duties at UNL include teaching technology students and developing a research program focused on advanced machinery systems for row crop production.