A couple of months ago, we revived an old project, the Hackerboat. The original project was building a boat to circumnavigate the globe autonomously. This was, to put it mildly, rather a bigger thing than we could chew. So the revival is a bit more modest in scope — we’re building an autonomous boat capable of doing sonar surveys of dive sites in the Puget Sound. Once that’s working, we’ll go from there. Since Toorcamp was then a couple of months out, we made that our target. The boat was there, but not entirely functional. More on that later.
We chose solar-electric propulsion as the easiest to implement quickly. Truly long distance cruising will require something like sail or wave power, but that’s immaterial to our more limited goals for operations within Puget Sound and in near coastal waters.
We bought an old kayak off Craigslist and set about modifying it. We made the following major modifications:
- Added center, forward, and aft bulkheads
- Added a keelson to allow us to bolt a keel and other equipment to the bottom of the boat
- Added stringers under the deck along the rails between the forward and aft bulkheads for mounting solar panels etc
- Built up the hatch comings to flat surfaces for simple hatches
- Motor mount for a trolling motor
- Geared, waterproof steering servo
- Lead-acid batteries
- A mounting frame for solar panels
- A radio mast
- Control electronics
The physical modifications ended up taking more time than planned, leaving the software and testing than would have been optimal. There was still significant amounts of electronics assembly, all the final integration work, and all but a tiny bit of the software to write when we left for Toorcamp with the boat in tow.
At Toorcamp, thanks to a tremendous amount of help from incidental volunteers, we were able to finish the electrical work, test the solar panels, test the low-level control electronics, test the motor, and add some status lights. We even got the radios up and working.
On Friday, we did a basic float test with an empty boat (no batteries, no solar panels). It floated well and we devised a system of ropes (thanks to Ratha and Trout) to make it easier to lift and carry the boat. There was no problem handling it in the low surf present that day. We also dropped the motor and connected the steering linkage in the waves. While not an ideal operation, we determined that this was a doable thing.
While we were able to get the low level control code up and running, we were not able to get the GPS navigation code or the shore station code running. With the end of Toorcamp fast approaching, we decided to launch anyways on Sunday and use the low level panic mode (run due east) to demonstrate the boat in action. We left off the keel, due to the difficult of beach launching with it in place. Neither of these turned out to be a great decision.
The lack of the keel is what truly doomed the attempt. We took the boat out into the surf, dropped the motor, and maneuvered it out to where Jenny was waiting for us. At that point, we had about ten minutes before the low level control system concluded that all was broken and it needed to run in to shore.
The weight of the boat more that doubled with the addition of the solar panels and the batteries. We re-arranged the ropes so eight people could carry it, and took it down to the water. When we took it out into the water, we discovered two problems. First, the front hatch was askew. We fixed that. Second, the boat was listing to the right. We re-arranged the batteries (while chest-deep in the surf, naturally) to take care of some of the list. Not enough, as it turns out. As soon as Jenny took the boat under tow and we let it go, it rolled over. We grabbed it, turned it back over, and hauled it out onto the beach, done for the day.
There was some water in the boat when we got it onto the beach, but not as much as I expected given that waves had broken over the askew hatch and we’d rolled it completely over. Nothing appeared to be broken, but the batteries were loose in their mounts and of course there was salt water on the outside of the electronics boxes.
Our current plan is to work to correct the mechanical, stability, and software deficiencies and get the boat back in the water at the end of August. We’ll do it in far more protected waters this time.