Indoor Autonomous Drone
A drone that navigates autonomously where GPS can't reach — using optical flow, camera vision, and custom embedded firmware.

## Problem
GPS doesn't work indoors. Warehouses, hospitals, disaster zones — the environments where autonomous drones could help most are exactly where satellite positioning fails. Indoor autonomy demands the drone perceive and localize itself using only onboard sensing.
## Solution
A drone that fuses optical flow, camera vision, and inertial data to estimate position and navigate autonomously in GPS-denied environments — with the perception and control loops running on embedded hardware onboard.
## Demo
## Features
- GPS-denied position estimation via optical flow
- Camera-based perception for navigation
- Sensor fusion of vision + inertial measurements
- Custom embedded firmware for real-time control
- Autonomous flight without external tracking systems
## Tech Stack
## Challenges
- Drift accumulation in optical-flow position estimates
- Real-time constraints on embedded compute
- Tuning control loops for stable indoor flight
## Lessons Learned
- One firmware stack, two worlds: optical flow indoors, GPS outdoors
- Supporting both autonomous and radio-controlled modes made flight testing survivable
## What's Next
Recognition: Best Demo Award at IEEE PIMRC 2023 in Toronto. Built as our University of Windsor capstone project: flight controller design, positioning, and obstacle avoidance.
## Gallery

