Heterogeneous Team of Robots: Sampling in aquatic environments

Alberto Quattrini Li, Dartmouth College

12/3/2019

Location: Upson 106 Conference Room Next to the Lounge

Time: 3:00p.m.

Abstract: How can robots effectively explore, monitor, and sample in large scale aquatic environments? This talk presents a recent interdisciplinary project funded by the National Science Foundation on monitoring cyanobacterial blooms in lakes with a team of heterogeneous robots. I will present a sample of solutions that involve the development and deployment of aquatic robotic systems for data collection. First, I show our efficient multirobot algorithms for a team of Autonomous Surface Vehicles governed by Dubins vehicle dynamics to cover of large areas of interest. Field trials with custom-modified motorized kayak are presented, providing insights for improvements.

Second, I discuss the use of a heterogeneous team of robots to exploit their complementary capabilities to reduce the operational cost and increase the mission time for environmental monitoring and water sampling. Using machine learning techniques to model the distribution of the observed phenomena, we developed adaptive exploration and sampling strategies that accounts for reduction in uncertainty. Experimental results from several field experiments together with some lessons learned will be presented.

The talk will conclude with a discussion on some of the open problems that still need to be fully addressed for a robust multirobot system useful for addressing environmental problems and current work, such as ensuring high-quality data and recovery mechanisms, towards the long-term goal of a ubiquitous collaborative multiagent/multirobot system for accomplishing large scale real world tasks.

Bio: Alberto Quattrini Li is an assistant professor in the Department of Computer Science at Dartmouth College and co-director of the Dartmouth Reality and Robotics Lab. He was a postdoctoral fellow and research assistant professor in the Autonomous Field Robotics Laboratory (AFRL), led by Professor Ioannis Rekleitis, in University of South Carolina from 2015 to 2018. During 2014, he was a visiting PhD student in the Robotic Sensor Networks Lab, directed by Professor Volkan Isler, at the Department of Computer Science and Engineering, University of Minnesota. He received a M.Sc. (2011) and a Ph.D. (2015) in Computer Science and Engineering from Politecnico di Milano, working with Professor Francesco Amigoni. His main research (currently funded by the National Science Foundation) include autonomous mobile robotics and active perception, applied to the aquatic domain, dealing with problems that span from multirobot exploration and coverage to multisensor fusion based state estimation. He has worked with many ground and marine robots, including Autonomous Surface Vehicles and Autonomous Underwater Vehicles.