Date of Award

Summer 2023

Access Type

Dissertation - Open Access

Degree Name

Doctor of Philosophy in Electrical Engineering & Computer Science


Electrical Engineering and Computer Science

Committee Chair

Radu F. Babiceanu

First Committee Member

Eduardo Rojas

Second Committee Member

Omar Ochoa

Third Committee Member

Bryan Watson

Fourth Committee Member

Nabeel Yousef

College Dean

James W. Gregory


Teams of low-cost Unmanned Aerial Vehicles (UAVs) have gained acceptance as an alternative for cooperatively searching and surveilling terrains. These UAVs are assembled with low-reliability components, so unit failures are possible. Losing UAVs to failures decreases the team's coverage efficiency and impacts communication, given that UAVs are also communication nodes. Such is the case of a Flying Ad Hoc Network (FANET), where the failure of a communication node may isolate segments of the network covering several nodes. The main goal of this study is to develop a resilience model that would allow us to analyze the effects of individual UAV failures on the team's performance to improve the team's resilience. The proposed solution models and simulates the UAV team using Agent-Based Modeling and Simulation. UAVs are modeled as autonomous agents, and the searched terrain as a two-dimensional M x N grid. Communication between agents permits having the exact data on the transit and occupation of all cells in real time. Such communication allows the UAV agents to estimate the best alternatives to move within the grid and know the exact number of all agents' visits to the cells. Each UAV is simulated as a hobbyist, fixed-wing airplane equipped with a generic set of actuators and a generic controller. Individual UAV failures are simulated following reliability Fault Trees. Each affected UAV is disabled and eliminated from the pool of active units. After each unit failure, the system generates a new topology. It produces a set of minimum-distance trees for each node (UAV) in the grid. The new trees will thus depict the rearrangement links as required after a node failure or if changes occur in the topology due to node movement. The model should generate parameters such as the number and location of compromised nodes, performance before and after the failure, and the estimated time of restitution needed to model the team's resilience. The study addresses three research goals: identifying appropriate tools for modeling UAV scenarios, developing a model for assessing UAVs team resilience that overcomes previous studies' limitations, and testing the model through multiple simulations. The study fills a gap in the literature as previous studies focus on system communication disruptions (i.e., node failures) without considering UAV unit reliability. This consideration becomes critical as using small, low-cost units prone to failure becomes widespread.