We installed a number of fairly standard solar mounting rails on BaanLoiNaam last week:

The goal is to have space for at least 12x 300W solar panels. Initially, I’m only planning on installing 8 of them, which should yield approximately 12-14 kWh of electricity per day on average. This would be more than enough for typical home usage, let alone the energy efficient small house design we have.

But the catch is that this house is meant to be autonomously floating in the pond. So every breeze or gust of wind will blow the house and the house needs to drive its propellers to keep itself centered in the pond and not crash into the edges of the pond. These propellers need power to run, and the solar panels are the only power the house has.

Pommm recently found me a new supplier of solar panels that are significantly cheaper. We’ve already noticed a significant markup in the prices of these panels in Thailand compared to the price if we ordered directly from China. But this vendor seems to have a much more reasonable markup. The prices drop from about $280 per panel (where we previously purchased them) to about $160 per panel for these 300W panels. I’ll recheck the prices from China, but if they haven’t gone down so much, I won’t need to worry about ordering directly from China anymore.

The other kinda crazy idea I have is that when the house has no other directives, it can move around in the pond and find the sunniest spots and then rotate the house so the panels are pointing towards the sun as much as possible. Why mount complicated solar tracking units for the panels when the whole house can turn? 😉

The intention is to do this during the day when we are not home as well as any other times we don’t really care what direction the house is pointing. It might get us 10-15% more solar energy from the same panels. But we’ll have to see about that.

All of this depends on getting a reliable control system that won’t go nuts and drive the house into the edge of the pond when we aren’t around. Reliability is usually the hardest part of engineering.

Based on personal experience, it is much easier to build a robot that can autonomously kick a ball into a goal when you have a team of grad students to set it up, make sure that batteries are charged, calibrate the sensors and double check the wiring before hand. (We used to compete in the international humanoid Robocup robot soccer competitions.) But to build a robot that anyone can flick a switch to turn it on and it stands up and reliably identifies the ball and goal without any calibration, the wires never come loose, etc… well this is probably about 10x harder. And for something as big as an autonomously driven house to run unattended without an engineer around to pull the plug if something is going wrong, that is going to require a level of confidence in the engineering more than anything I’ve done before.

So I’m going to try to build it, but I’m not 100% confident of success. It will be up to passing a set of simulated stress tests under supervision reliably and repeatably before I would even consider letting it go unattended.