WaterFeeler2 goes to sleep due to constant bead weather but fulfills mission goals:

SUCCESS

It's official. WaterFeeler2 has completed its primary mission goals. We are now in an extended mission. That's the good news.

The bad news is that we are currently suffering from low power conditions due to constant bad weather with rain and heavy overcast. The probe still powers up during the day but goes to a shutdown during the night.

Low power caused by bad weather

Let's take a look at the voltage graph first:

As you can see, there is pretty much no recharging happening. We just get enough energy to run the Arduino and the RF24 radio durin the day. If we take a look at the weather conditions we see why:

Basically, it has been raining for four days. Maybe, if i would have been much more agressive in designing the electronics with powersaving in mind, we could have weathered those conditions. But as it is, i forgot to disable some power leds and the Raspberry Pi/PanCam voltage regulator is on even when the power to the Raspberry itself is off.

And, oh yes, the cheap solar charger is only PWM and does not do MPPT (maximum power point tracking), so we are loosing quite a bit of potential solar power due to that. But there aren't that many affordable MPPT chargers for 6V systems out there. For the next project, i'm probably switching to a more classic 12V off the shelf system - it probably still will not be MPPT due to costs, but you can get reasonably cheap solar panels of a decent size/power for those systems, so just cranking up the panel size might be the cheaper way to go here.

I'm also thinking of designing the next probe electronics (or at least the core part) on a PCB instead of a breadboard/perfboard. This should make it way easier to coat the whole thing in a clearcoat to make it more resistant to condensation (as well as reuable for follow-up projects). Just look at the humidity data:

We are pretty much at the upper third of the graph here and we were even hitting the 100% internally. Which means there was condensation. But so far, the electronics keep working. But i'm worried about long term corrosion.

New images

High resolution single frame

I've got a couple of new images to show. The first one is a high resolution single frame image, looking out from the porch. It clearly shows raindrops on the mirror and up on the solar panel. A bit harder to see is the standing water in the RFI antenna (the pink thing with the yellow wire around). It's not up to were the wire comes around to the foreground the first time (so probably less than a centimeter), but we DID collect some rainwater. Yay!

High resolution panorama

This is the panorama of pretty much the same day (taken a minute or two after the above single frame). The light has changed a bit, probably due to a cloud moving in front of the sun. Apart from a wider (higher?) view, one important thing it shows us is that the solar panels are wet, but reasonably clean.

There also doesn't seems to be water pooling in the bottom of the electronics box (the one blob is hot glue, if i remember correctly).

Bad color resolution

You might have noticed on the images above that the color resolution is not the greatest and there is a lot of dithering going on. Part of that is due to image resizing and using and old Raspberry Pi "Version 1" camera.

But i also do suspect that there is something wonky going on with my image processing. Once we get some nice, sunny weather (so we have ample power), i'll see about capturing and downloading some raw images to look into that some more.

I have plotted the color data of the above panorama on a graph:

Clearly, only certain brightness values for each color get used, meaning we don't seem to get 8 bit data for each color channel. This certainly looks like an image processing snafu.

Here are the number of "channels" used for each color and the total:

So far, it looks like we are getting only 7 bit resolution per color channel. On the bright side, it's not the problem i feared i was getting, that is accidently using an "indexed" format during processing, which would map the RGB colors to 256 possible color. But still, there is something wonky going on, and i need to investigate more.

Mission status update

Let's take a look at our mission objectives and the status of each one. We got all of the "Required" objectives (although we need to redesign stuff for future missions). Many of the "Desired" ones are checked, the rest is in progress. We even fulfilled three of our "Optional" objectives.

Required

These are the objectives that must the fulfilled. If anyone of these don't work, it's a mission failure.

Desired

These are the nice-to-have objectives. If they don't work or can't be implemented in time, it's sad but not critical. But they must not interfer with the fullfilment of the "required" objectives.

Optional

These optional objectives are neither urgent nor terribly important. But they could accelerate the Garden Space Program by providing additional data points and early test results.

tl;dr

"Bad weather is bad for solar power. Pretty images with some color processing flaws. Mission success!"