World of Drones
A camera drone that spots vehicles from the air and reports them to a live dashboard.

An end-to-end IoT system built during the GlobalLogic summer internship. A drone carries a Raspberry Pi with a camera and runs a lightweight object-detection model on board. It marks vehicles and their positions during the flight. A FastAPI backend collects the detections, and the DroneVision dashboard in Next.js shows the flight path, the recorded video and the details of each vehicle.
Backstory
World of Drones was the project of the GlobalLogic IoT project-based learning programme in the summer of 2024: a few weeks of Scrum sprints, code reviews and a final showcase in front of the company. My first job was the enclosure that mounts on the drone. In Autodesk Fusion I designed a 3D-printed case that holds the Raspberry Pi, the STM32, a LiDAR, a GPS module, an accelerometer, a gyroscope and the camera. Everything had to fit within the drone's limits on weight and space. When the enclosure was done, I moved to the software and built the whole frontend of DroneVision, the operator dashboard. I designed the UI first, then wrote it by hand in Next.js: the overview, video playback, the detections table and the flight map with the details of each vehicle. At the showcase we demonstrated live vehicle detection from the drone.
Highlights
Sensor Enclosure
Raspberry Pi, STM32, LiDAR, GPS, IMU and camera in one 3D-printed case
DroneVision Dashboard
Designed first, then hand-written in Next.js
Flight Map & Detections
Flight path, video playback and the details of each vehicle
On-Board Vehicle Detection
A lightweight vision model running on the drone itself
Agile Delivery
Scrum sprints, code reviews and a final showcase at GlobalLogic