Maximum Inertia

Introduction

Our goal was full integration with the Maxim board and the flexibility to adapt the software to many different handyman-style use cases, e.g. as a table level, compass, etc. We experimented with and implemented various noise filtering techniques such as Kalman filters and low pass filters to increase the accuracy of our data. In general, working with the Pegasus board involved a lot of low-level read-write operations within internal device registers, so basic tasks like getting accelerometer data became much more complex than we were used to.

The sensors consist of the Maxim Pegasus board and any Android phone with our app installed. The two are synchronized at the beginning, and then by moving the "tape" away from the "measure," we can get an accurate measure of distance, even for non-linear surfaces.

Challenges

Sometimes high-variance outputs can come out of the sensors we made use of, such as Android gyroscopes. Maintaining an inertial reference frame from our board to the ground as it was rotated proved very difficult and required the use of quaternion rotational transforms. Using the Maxim Pegasus board was difficult as it is a relatively new piece of hardware, and thus, no APIs or libraries have been written for basic functions yet. We had to query for accelerometer and gyro data manually from internal IMU registers with I2C.