Flipped In The Surf

Usually every Thanksgiving, we drive down in the van to visit my parents at their cottage in Cap-Brulé. This weekend was no exception, and of course AMOS was squeezed into the van along with 5 people, a dog, and my kayak on the roof. One of my parents' neighbors had wondered if AMOS would be able to measure the distribution of water temperatures along the coastline where a lagoon empties into the ocean, so I was hoping to give it a try. The first day we were there was quite windy and wavy, with rain and gusts up to
40 km/hr, so AMOS stayed in the shed. The following day was better in the afternoon, so my mother, father, and I drove it down to a small boat launch area near the cottage.

The waves were still a bit too strong to successfully launch AMOS from shore, so I pulled it along in the kayak for a while, and then let it go on its own, following the planned GPS sample points. At one point, when it was on its own though, fairly close to shore, a strong wave struck it broadside and flipped it over solar panel down. The propeller post snapped off on the sandy bottom and the electronic control boxes were both submerged, under the boat for about a minute before I was able to frantically paddle over to set things right-side up. 

We didn't want to get salt water from AMOS or the kayak on the van, so Dad (who is 77) and Mom (73) both carried AMOS approximately 1 km back to the cottage, while I carried the kayak. They got there first, so Dad then came back to help me finish carrying back the kayak. I then had a look at the damage:

The main processor box only had a few drips of water in it, but the battery box had a few mm of water resting on its bottom. Probably some of the cable glands could have been a bit tighter. While manipulating the battery and wires, the +12 V contact shorted out on another wire with a loud pop and some smoke. Uh oh. Turns out it was only a digital I/O from the Raspberry Pi that got fried though, so no big deal. I was able to just change the software to use a different one. And actually, although the flip prevented me from collecting about half of the data that I was hoping for, the plots of temperature and turbidity still turned out pretty good (https://www.innaturerobotics.com/sample-data-1):

The water intrusion didn't seem to cause any harm inside the battery box, but so far I haven't been able to get the camera module working. There were still a few drops of condensation inside the plexiglass camera cover, so I'm hoping the newly re-charged desiccant that I put in there will dry things out enough so that it works again soon. 

I guess for wavy conditions, it might make sense to have a keel, or at least some fins on the bottom of AMOS to help stabilize it. Maybe an add-on option that customers can purchase. ☺