Well, I Don't Think This Will Work For Me

[whitedavidp]

I have spent at least a week now trying everything I can (settings, photo adjustments, different models, different camera positions, etc) to get consistent readings of my ancient water meter located in a small crypt by the street. I planned to continue the battle, perhaps even trying this alternative reading mechanism (which would NOT really require On-The-Edge).

But the physical situation just seems to add an insurmountable set of problems regardless of the software:

1. Even though located below grade and fairly well covered, there is just too much moisture in the crypt and it condenses both inside and outside the clear cover on the meter. This makes it highly unlikely I will ever be consistently able to capture high quality images over time here in the Pacific Northwest.
   
2. Being located below grade with an exposed, dirt bottom,
   
   the crypt is endlessly pestered by insects and slugs/snails. At times, they crawl over the surface obviously obscuring images and often leaving trails that persist after their passing.
   
   I considered adding cayenne pepper, diatomaceous earth, crushed oyster shell, etc to the crypt to repel these creatures (who obviously have a right to be there). But most of these loose their effectiveness over time and I really do not need more things to maintain around here.
3. Out at the street, far from any utility A/C outlets, providing power is a problem. Even with this amazingly efficient mechanism (I found that, even with all the work it is doing on-board, the power consumption of this device and firmware is about the same as just having Tasmota installed, running - but not doing much of anything), providing power is not easy. I originally thought I would use a battery pack of some sort and put solar cells up to keep the battery topped off. But that is a whole new rabbit hole of unknowns to me with success far from guaranteed. And, my original intent was to employ - to the extent possible - bits and pieces of stuff I had otherwise just lying around here (read: on the cheap). As a proof of concept, I used a solar charging, lithium iron powered work light that I was keeping on my car's rear deck.
   
   But the limited exposure to the sun for this small (5 watt, I think), single solar cell was simply not sufficient.
4. I know that many/most users of On-The-Edge either have their meters in more benevolent locations OR are able to contain the meter's surface with mounts that might prevent (1) and (2) above. But I cannot do this. Our water company will insist on its periodic visual reading of the meter for purposes of creating bills (then are far too backwards technically and otherwise to accept a printed photo from me as proof of the meter reading). So the meter's face itself needs to be left un-obscured and easily read. Hell, it is unlikely the clods they employ for meter reading would be able to read the meter anyhow without jarring my camera setup anyhow.

On the "success" side, I did manage to get reasonable wifi out to the street and the ESP32 camera using an external antenna I had lying around the house already. I had to get help from a much more talented friend with the micro-soldering required to switch from the pretty lousy on-board antenna. And I learned that not all antenna connections are the same - the central conductor in the co-ax may or may not be present on either the male OR female side of the connection (I ended up with both male and female sides of the bits I had on hand lacking this central conductor making no antenna at all - I "solved" this by snipping a piece of sewing needle and pressing it into one side).

So this has been an interesting, even compelling, project for me to pass some time. I have come to like the simplicity of the ESP32 camera and to respect its ability to perform significant tasks on its own (in the right hands). And I will, no doubt, re-deploy the units I have for other, more traditional surveillance/security work around the house. So none of this has been a waste.

I leave this simply as a record that others may find useful. Best!

[SybexX]

https://www.thingiverse.com/search?q=Ai+On+The+Edge&page=1

[SybexX]

I developed my own housing and printed it with the 3D printer.

[whitedavidp]

I see. No pun intended.

If there is moisture under the meter's cover, I wonder if an external housing as you describe would be enough. Also, how to permit ham-handed meter reader folk to see/read the meter? If the housing were very clear/transparent, that might work. But I doubt that would be possible with a 3d printed material. Of course, I know nothing about 3d printing so maybe my impression is incorrect.

Thanks for all of your help and encouragement.

Well, I just did some googling and it does seem that transparent 3d printed parts are indeed possible.

[SybexX]

The housing is aligned through the lid so that you can easily remove it and then add it again.

[SybexX]

Or alignment by magnets https://www.thingiverse.com/thing:5909513

[whitedavidp]

Thanks for all. I like the magnetic alignment trick. All along here, I have been thinking that the water system employees would have to remove such a covering in order to read the meter. And I was (and remain) pretty sure that they would either complain or mess things up - perhaps breaking things entirely. I typically know in advance which days are scheduled for meter reading and it seems possible I might be able to remove things off to the side of the crypt on those days. Rather more than I was hoping to have to do to maintain this. But it is possible, I suppose. And your suggestions would certainly make sense in that case.

Even if I do all this, I will still have to figure out a way to get more consistently correct readings here. I wonder if this other watermeter system (not on-the-edge but linux-based) might do better for my ancient meter. Perhaps I can try it out using On-The-Edge as an image provider? I have it installed on my H/A server but never explored it after finding On-The-Edge.

Best!

[whitedavidp]

Well, perhaps I over-reacted. In the last 24 hours, things have at least dried out substantially.

I need to consider how I hope to use this mechanism. My #1 goal is to identify leak conditions - water use beyond normal expectations - so I can send alerts/alarms from my H/A system. I have a friend who just got an $8000 water bill due to an undetected broken pipe between the meter and his home!

It would be great if I could do more in the manner of water use analysis and conservation based upon this data. But that is very secondary. So if I can manage to get accurate readings 2-4 times in a 24 hour period, that might be all I need and perhaps I shouldn't be so fixated?

Best.