Once our system was setup, we went to collect data.
The Frisbee was held ready, and then the power was turned on, beginning data collection. We laid it flat for a few seconds, first, and then threw it. After the Frisbee was caught, we turned off the power, stopping data collection. The beginning period of stillness was used later in data pre-processing.
In order to help validate our model, we took a video of the throw, and also recorded the distance it traveled.