L2M – AutoSAR: Autonomous UAV Search and Rescue Using Deep Learning for Emergency Response in Remote and Infrastructure-Free Regions.

In remote and rural regions, where communication infrastructure is sparse and GPS is unreliable, emergency Search and Rescue (SAR) operations should be fast, reliable, and cost-effective. Existing technologies, such as thermal imaging and mobile-based systems like Cellular Airborne Sensor for Search and Rescue (CASSAR), depend on costly sensors, human interpretation, or the presence of a mobile device, which limits their effectiveness in many real-world scenarios. The AutoSAR solution is designed to fill this critical gap. AutoSAR is an autonomous UAV-based search and rescue system for emergency response in infrastructure-free regions. The UAV emits and receives its own signals to detect and localize individuals in distress, eliminating reliance on GPS, networks, or external sensors. By leveraging deep learning, AutoSAR estimates position and velocity from reflected signals in real time, enabling rapid, intelligent rescue missions in rural areas such as mining sites, wilderness trails, and northern communities. This technology addresses a national public safety need by offering a scalable, cost-effective, infrastructure-independent SAR solution.

Faculty Supervisor:

Salama Ikki

Student:

Partner:

DMZ Ventures Inc

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

Lakehead University

Program: