Figure 1: AI enabled camera system performing real time inference during drone flight
By Frederick Andrews, Jad Mghabghab, Mouad Ben lahbib, Maureen Kouassi, Josué Dazogbo, Computer Engineering Students at the University of Ottawa
Date: 3 July 2025
Figure 2: First field day group picture
T.A.I.L.S. (Tactical Aerial Insight and Localization Suite) is a drone-based Point-of-Interest (POI) mapping solution designed to enhance search and rescue operations, wildlife monitoring, and coastal surveillance.
The system integrates AI-powered image recognition with GPS tracking to detect and mark important locations on an interactive map within a mobile application. This project involves expertise in artificial intelligence, real-time embedded programming, wireless communication, and network security.
Figure 3: Assembled camera system mounted on top of a drone
- 🧠 AI Object Detection & POI Recognition: Detect and classify objects of interest and transmit them as Points of Interest (POIs).
- 🛰️ Long Distance Telemetry Transmission: Real-time, long-range communication of drone position and altitude.
- 📱 Mobile App Integration: View past flight results, positions of recorded data points, and historical POIs directly in the app.
The onboard module features a Raspberry Pi 5, battery pack, LoRa-GPS module, and a Hailo AI hat for accelerated inference. It uses a Raspberry Pi Camera HQ with a wide-angle lens to capture images, reads GPS signals from sensors, and creates LoRa packets for transmission.
Figure 4: Custom adjustable mount for Raspberry Pi High Quality Camera
The base station utilizes a modified ESP32 named Heltec LoRa board supporting both BLE and LoRa communication. It receives LoRa packets from the drone and converts them into BLE packets to send to connected devices.
Figure 5: Base station converting LoRa packet to appropriate BLE packet
The mobile application is built with React Native using Expo Go and Firebase for backend services. It displays Points of Interest (POIs) on a map, along with flight data, and is accessible on both mobile devices and the web.
Figure 6: Mobile app developped for POI lookup and flight telemetry
Below is a detailed list of hardware components used for the TAILS Embedded System, with all costs shown after tax.
| Item Description | Quantity | Cost After Tax ($) |
|---|---|---|
| Raspberry Pi Zero 2 W + AI Camera + Accessories | 1 | 240.01 |
| Raspberry Pi 5 (16GB) + Accessories | 1 | 304.30 |
| Raspberry Pi HQ Camera + Lens | 1 | 137.24 |
| USB to LoRa Dongle | 1 | Included Above |
| Heltec LoRa Board (Base Station) | 1 | 41.69 |
| SX1262 LoRa Hat | 1 | 52.93 |
| AI Accelerator Hat (Hailo-8L) | 1 | 123.06 |
| PHAT-GPM GPS Module | 1 | 55.99 |
| Raspberry Pi UPS Power HAT | 1 | 77.79 |
| Adjustable Camera Mount | 1 | 35.65 |
| Drone Parts (Motors, Frame, ESC, etc.) | — | 754.77 |
| Accessories (Cables, spacers, mounts, etc.) | — | 159.67 |
| Drone Assembly Fee (StansUAV) | 1 | 339.00 |
| Total Project Cost | — | 2,322.10 |
