Photo credit : Lawrence Taylor
Later that day Lawrence and I drove to Whycocomagh in Cape Breton where we had rented a cottage on the shore of the Bras d'Or Lake. Here was our view from the front porch the next morning:
During our stay in Whycocomagh, we were hosted by Bruce Hatcher, the University Chair in Marine Ecosystem Research at UCCB and the director of the Bras d'Or Institute. He had a Zodiac boat:
Photo credit: Lawrence Taylor
and I introduced AMOS to them, before setting it off on its pre-programmed course:
Video credit: Lawrence Taylor
Things were going well, just as before:
Video credit: Lawrence Taylor
but then AMOS unexpectedly disappeared near a large grey boat that was parked at the corner of the opposing dock. Bruce and a couple of his students went over to investigate in the Zodiac. When they approached the area, one of the students exclaimed "Oh shit Bruce, I think it must have gone under the dock and disintegrated!" Upon closer inspection however, AMOS was found underneath the large grey boat, which actually happened to be a catamaran. They were able to retrieve it, and since it did not seem to be moving anywhere on its own, brought it back to our dock, where it was discovered that the propeller blades had both been broken, most likely when it was underneath the catamaran. Unfortunately, I had left the spare propeller back at the cottage, so that was the end of the demo. AMOS did capture some interesting footage from its experience though:
After the demo, Lawrence and I went out with Bruce and Allison and Piotr to check out a 50 m "Death Hole" with a couple of ROVs. This hole featured a mysterious layer of floating material about 16 m deep that both ROVs were able to examine with their camera systems. Allison and Piotr's ROV collected water samples for their lab at Dalhousie University using a very cool apparatus of spring actuated syringes. Bruce's ROV collected a smelly, dark sample of mud from the bottom of the hole:
On the return trip back to Dartmouth, Lawrence and I stopped near Antigonish to try collecting some data in Pomquet Cove. The wind here proved to be a bit too much for AMOS, so we tried a stream on the other side of the highway instead:
Video credit: Lawrence Taylor
The water in the stream was a bit mucky and shallow in some areas, and required a few minutes of manually controlling AMOS to get it unstuck and drive it back to a location where it could be easily retrieved.
Overall, the demos were a good experience, and served to highlight a number of key areas that will require further development:
- More power, need a larger propeller, larger motor, and larger / more batteries.
- More streamlined and durable construction.
- More accurate positioning (i.e. requires a GNSS system).
- Need to have ability to reverse direction of propeller for backing out of tight spots.
- Software improvements for defining survey grids, saving / loading standard GPS coordinate file formats.
Thank you for this prompt follow-up Murray!
ReplyDeleteApparently you learned a great deal from the sea-trials of the AMOS, and it was a good experience for us all to work on a functioning fish farm in the Bras d'Or Lake Biosphere. For me, the great strength of your invention is the AI algorithm you developed: helping the AMOS to think like a Captain in the perilous waters of the coastal environment.
I look forward to round two of the Cape Breton match-up!
Best, Bruce
Thanks Bruce! I don't know if AMOS can really think like a captain yet; there's probably still some room for improvement there! :-)
ReplyDelete