CPS: Small: NSF-DST: Autonomous Operations of Multi-UAV Uncrewed Aerial Systems using Onboard Sensing to Monitor and Track Natural Disaster Events

This research project focuses on using uncrewed aerial systems (UAS) to monitor and track natural disaster events like wildland fires and flooding rivers and lakes. As the intensities of these events continue to increase worldwide due to climate change, the need for their early monitoring and tracking has also increased. A UAS, consisting of a team of uncrewed aerial vehicles (UAVs) and one or more ground stations, can provide real-time monitoring and tracking of unfolding disaster events, as well as help with relief operations to avoid large scale losses of lives and property. This project brings together US and Indian experts working on autonomous UAS operations in the presence of environmental uncertainties and hazards. In particular, it seeks to understand how teams of autonomous UAVs could be used to maximize the data gathered and predict the intensity and spread of forest fires and floods. Through its outreach activities, this project team will also encourage students to participate in research on autonomous vehicles and inform the general public about the value of such research in addressing societal challenges. Key research goals are to design and use nonlinearly stable and robust motion estimation and control schemes that enable a multi-UAV team to collaboratively follow desired trajectories for loitering, monitoring and tracking in the presence of disturbances like wind and air currents. Each UAV in the UAS is modeled as an actuated rigid body, making the UAS a multi-agent rigid body system (MARBS). Geometric controller and observer designs will be developed that are computationally light and can be implemented with commercially available sensors onboard rotorcraft UAVs. Sensor data from inertial sensors and point cloud sensors like depth cameras, will be combined using a continuous and finite-time stable extended state observer (ESO). This ESO will provide estimates of UAV motion states, relative pose of other UAVs within the range of point cloud sensors, and disturbance estimates. These estimates will be used by feedback tracking controllers that are designed to track desired trajectories in a stable manner while rejecting the disturbances, thereby providing active disturbance rejection control (ADRC). Both asymptotically stable and finite-time stable control laws will be designed, and the entire navigation and control system for a multi-UAV team will be tested in indoor and outdoor environments. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.