Computer Science Seminar by Ankit Agrawal

Speaker: Ankit Agrawal, assistant professor of computer science, Saint Louis University 

Abstract: 

The rapid integration of artificial intelligence (AI) into unmanned aerial vehicles (UAVs) has unlocked transformative potential for critical real-world applications, including river search and rescue, city surveillance, and wildfire monitoring. However, deploying AI-driven systems in unpredictable, dynamic environments introduces complex challenges. For instance, imagine a UAV navigating an urban area with tall buildings, such as Chicago Downtown, during a fire emergency: turbulent wind conditions combined with intermittent GPS signals could cause the drone to deviate from its planned course, risking a crash or delayed response. Such scenarios highlight the critical need for robust systems that operate flawlessly even under extreme conditions to ensure mission success and prevent catastrophic failures. In this presentation, I will discuss my research, which adopts a software engineering perspective to advance three key areas: (1) novel AI-based simulation testing methodology to rigorously evaluate the safety and reliability of AI-driven UAVs; (2) a high-fidelity test-bed that simulates urban environmental factors primarily responsible for drone crashes, such as turbulent wind conditions and GPS signal uncertainty in order to enable the early identification of safety risks; and (3) system design innovations, including augmented reality (AR) interfaces, to enhance situational awareness of end users, such as police officers and firefighters, who rely on UAVs as mission-critical tools in high-stakes operations. 

Bio: 

Ankit Agrawal is an assistant professor of computer science at Saint Louis University. He received his Ph.D. in Computer Science from the University of Notre Dame, Indiana, in 2022. Ankit has published his work in top-tier software engineering conferences, including ICSE, FSE, and ASE, and his research has garnered accolades at premier human-computer interaction venues such as CHI. His current research, supported by NASA, focuses on advancing our capabilities to test modern AI-driven smart unmanned aerial vehicles. With a practical approach to both teaching and research, he collaborates with industry partners, contributing to initiatives such as DroneResponse, a spin-off that designs software autopilots for emergency response operations. Looking ahead, Ankit aims to further integrate emerging AI methodologies and interdisciplinary approaches to enhance the robustness and scalability of UAV systems, ultimately paving the way for their broader adoption in mission-critical applications.