Sensor Fusion Consistency Inquiry

[bkris]

Dear reader,

Context: We are currently using BNO055 in our products to track cyclical motion in space. There are also cyclical shocks on the order of 1g, in the cyclical motion. Currently, we are using the gravity vector to keep track of our orientation and we get drift sometimes in our orientation data from the IMU.

Inquiry: There are (at least) three options to get orientation information from the BNO055, gravity vector, Euler angles, and quaternions. Does the sensor fusion return the same orientation for these three quantities in each time step? Does each orientation metric have its unique pros and cons? Do the previous answers change if the IMU was BHI260AP?

Thanks,

bkris

[BSTRobin]

Hi bkris,

Could we know more information about your application first:
1. Which operation mode you configured for BNO055?
2. Could we know more detailed information about cyclical motion in space?

3. BNO055 was not recommended for new design, would like your design a new project with BNO055?

[bkris]

Hi BSTRobin,

Thank you for your reply.

1. The BNO055 is configured to 9DOF Sensor Fusion and we are using the gravity vector.

2. The BNO055 is used to track the orientation of a lower limb - i.e. walking.

3. We are not making a new design or a product. We are investigating why this has been an issue in parallel with evaluating a replacement for the BNO055. The objective is to minimize the gyro drift we observe from the IMU.

Regards,

bkris

[BSTRobin]

Hi bkris,

The outputs of gravity vector, Euler angles, and squares correspond to the same coordinate system. Euler angles, and squares are generally used in application.