Drone Position Hold using Optical Flow Sensor – Joystick Control with Dead Reckoning

This advanced undergraduate-level project implements a closed-loop position hold and joystick control system for the Litewing ESP32 drone using a LiteWing Flight Positioning Module. At the core of the system is an optical flow sensor, which optically tracks surface textures beneath the drone to calculate translational velocity and relative movement. Because raw optical flow measurements can drift or accumulate error over time, dead reckoning algorithms are integrated alongside cascaded PID control loops to maintain stable spatial coordinates. The ESP32 DevKit v1 processes sensor data and executes control algorithms in real time, communicating bi-directionally with a ground control PC running Python and the cflib library. When a user inputs joystick commands, the software translates those inputs into desired velocity or position setpoints. The onboard firmware then dynamically adjusts motor commands via the Litewing ESP32 drone's electronic speed controllers to counteract environmental disturbances like air currents, allowing the aircraft to maintain a stable hover or execute precise manual positioning maneuvers in indoor, GPS-denied environments.