Work / Electronics / Embedded / IMU Sensor Fusion
Signal Processing · Embedded
IMU Sensor Fusion
Attitude estimation system fusing 9-DOF IMU data (accelerometer, gyroscope, magnetometer) using a complementary filter and extended Kalman filter on a microcontroller.
Section
Electronics / Embedded
Stack
C · MicroPython · Kalman Filter · IMU
Category
Signal Processing · Embedded
Status
Complementary filter stable, EKF tuning ongoing
01 — Challenge
The core engineering challenge
Each IMU axis has different noise characteristics and drift rates — naive sensor fusion degrades over time in ways that only become visible in extended flight tests.
03 — Decisions
Engineering decisions and what they cost
01Prototype filter logic in MicroPython first, then port the validated algorithm to C.
02Empirical noise characterisation over 30-minute static sessions rather than relying on datasheet values.
03Log both raw and filtered output continuously so filter response can be visualised and tuned.
04 — Stack
CMicroPythonKalman FilterIMUDSPRaspberry Pi Pico