The pool trials with simultaneous use of two underwater robots went pretty smoothly, despite major last-minute roadblocks.
The night before the pool trial we discovered that a custom-made video-to-fiber-optic transducer board was not working. Chris tracked it down to a bad integrated circuit. He managed to desolder the surface-mount and order a replacement from digikey who expressed it to use the next day. It was soldered back in place in tiem for the pool trial.
Then, just a few hours before the pool trial we found a problem with one of the IMU's. This was tracked to a bad solder joint, which may be been undetected since a leak many months ago. That got fixed too and we made it to the pool trial just a little bit late.
The experiements went well and we did everything on the "high priority" list. We always have alist of extra experiments to do it there is time, but not many of these got performed. We did manage to validate several different model-based controllers for the robot. They seem to have done really well. We also collected various kinds of vision system data for training and testing, and validating human interaction experiments.
