Autonomous SAR Hexacopter
CLIENT
Self-Funded (IUB Senior Design)
YEAR
2026
INDUSTRY
Aerospace & Robotics
SERVICES
UAVs & Autonomy · ROS 2
Key Highlights
Onboard Edge AI Vision
Engineered a computer vision pipeline running locally on a Raspberry Pi 5. By fine-tuning a YOLOv8 Nano model and integrating an ESP32-CAM, the system achieves 25ms inference times (55.2% mAP) to detect survivors in disaster zones without relying on cloud processing or internet connectivity.
Redundant Communication & Autonomy
The hexacopter autonomously executes complex search patterns, such as expanding squares and parallel sweeps. The architecture includes a multi-medium fail-safe that automatically shifts to secondary communication links if the primary signal degrades, ensuring uninterrupted rescue operations.
1st Place IEEE Exhibition Winner
Designed and manufactured the entire autonomous system from scratch for under $630 USD, proving industrial-grade search and rescue capabilities can be achieved on a highly constrained budget. The system successfully executed an autonomous detection and payload drop during live testing.
Designed for measurable product growth
Every design decision focused on improving usability, increasing engagement.
25ms
Inference Latency
Per-frame processing speed achieved natively on the Pi 5.
4kg
Payload Capacity
Delivered life-saving equipment with a 3-meter drop accuracy.
$630
Total Build Cost
Engineered a robust hardware solution at a fraction of commercial alternatives.
