Gimbal Sensor Fusion Firmware Upgrade
Budget: $1,500 – $3,000 USD
I already have a working PCB built around an Xsens MTi-630 IMU and an NXP MKV56 microcontroller. What I need now is a technician who can revise the existing firmware so the remote camera head delivers rock-solid motion compensation on set.
Your task is to design and embed a custom sensor-fusion routine that blends gyro, accelerometer, and magnetometer data from the MTi-630, then feed that fused orientation into the motor-control loop over CANopen. I’m aiming for a balanced mix of speed and accuracy so operators never feel lag yet still enjoy drift-free performance during long takes.
You will be starting from my current codebase (MCUXpresso / NXP SDK), which already handles basic hardware abstraction and CANopen messaging. The focus is on:
• Writing or adapting a quaternion-based sensor-fusion algorithm optimised for the MKV56 (FPU enabled).
• Integrating it cleanly with my existing control system.
• Publishing the fused attitude, angular-rate residuals, and confidence metrics on the current CANopen PDOs.
Deliverables
1. Updated firmware source, project files, and build instructions.
2. Brief tuning guide so we can adjust filter gains for different payloads.
3. Reduce the drift by as much as possible in the current implementation.
If you have prior experience with Xsens devices, FreeRTOS on NXP parts, or CANopen profile CiA 402, please mention it—those skills will help you hit the ground running. I can supply the existing schematics, codebase, and a desktop test jig as soon as we start.
Your task is to design and embed a custom sensor-fusion routine that blends gyro, accelerometer, and magnetometer data from the MTi-630, then feed that fused orientation into the motor-control loop over CANopen. I’m aiming for a balanced mix of speed and accuracy so operators never feel lag yet still enjoy drift-free performance during long takes.
You will be starting from my current codebase (MCUXpresso / NXP SDK), which already handles basic hardware abstraction and CANopen messaging. The focus is on:
• Writing or adapting a quaternion-based sensor-fusion algorithm optimised for the MKV56 (FPU enabled).
• Integrating it cleanly with my existing control system.
• Publishing the fused attitude, angular-rate residuals, and confidence metrics on the current CANopen PDOs.
Deliverables
1. Updated firmware source, project files, and build instructions.
2. Brief tuning guide so we can adjust filter gains for different payloads.
3. Reduce the drift by as much as possible in the current implementation.
If you have prior experience with Xsens devices, FreeRTOS on NXP parts, or CANopen profile CiA 402, please mention it—those skills will help you hit the ground running. I can supply the existing schematics, codebase, and a desktop test jig as soon as we start.