Conclusion

Our system achieved our goals of weight and cost.  The system was hampered by two main factors: the choice of IMU and motors. In the inherent design of the system, the disturbance of the gimbal must be read as accurately as possible. The MPU-6050 performs well for a $5 IMU however the addition of a magnetometer (compass) would make the unit more reliable and would aid in further cutting down on noise. Upgrading the motors to brushless DC motors would also improve performance. While generally more expensive and heavier, they operate on a supplied voltage and therefore a more finely tuned controller could be used such as a PID or PV.

Future Considerations:

Aside from upgrading the IMU and motors to BLDCs, we would also work to:

  • Minimize free wires
  • Design the gimbal to endure wear

 

Add functionality such as:

  • Track a location or an object (allow for better photography or footage during a drone flight). This could be accomplished through computer vision or through GPS.
  • Live Video Feed: through the use of radio modules or WiFi, the feed from the Raspberry Pi could be transmitted live to provide better feedback to the operator.
  • Communication with the RC controller. In conjunction with a live video feed, button on the controller could then allow the operator to control the gimbal mid-flight.