Overall, we think our project went really well. The hat registered people’s names correctly, moved when it was supposed to move and was a lot of fun to use. Both visitors enjoyed talking to the hat and getting sorted. Sometimes the speech API would get confused by background noise, but worked really well in a quiet room.
If we had more time, we would like to make some improvements. When we started the project we assumed we could run the speech API on the MyRio, and later found out we could not. In order to get the hat to talk to you we had to have the speech code run on a computer. Instead of talking to the hat you had to talk to a screen. Fortunately, this didn’t take too much away from the project but was still a little disappointing. Also, we realized a little too late that having to click the screen to confirm your name was a bit of a hassle. We should have had that portion also recognize a yes or no answer. Lastly, it would have been nice to be able to provide more power to the motor. Since the MyRio and its battery are so massive, we didn’t have enough space to stuff in another board to connect the battery to the motors directly. The hat has a second motor that moved the mouth (very faintly) that we decided to ignore because the top movement was so much more impressive. If we had more time/different equipment, it would have been nice to use both.