With the weather getting colder, [Daniel] decided it would be a good idea to monitor how much energy his gas heating was using in real time. He used a Nokia 6680 cameraphone to monitor the heater’s flame through the sight glass. PyS60, a Symbian implementation of Python, checks the image sent by the camera and measures how much blue flame is visible. These values are stored in a SQL DB on the phone that can be polled over Bluetooth. At the end of the billing cycle, he’ll be able to correlate the amount of gas used with what the phone reported.
[Thanks, florent bayle]
Almost seems like overkill, though natural gas isn’t as cheap as it once was. Nice implementation.
On a personal note, I would do something like this but my place is heated via heating oil, which doesn’t leave much choice. I get a 110-gallon tank filled twice a year, and all I have to do is keep track of how much is left around December.
i cant imagine this is even remotely accurate. A better way would be to use a wind meter inline with the gas flow.
I hope for your sake that is not a Rinnai heater.
@rivetgeek
it’s accurate because the gas is burned at a constant rate. the camera doesn’t analyze the properties of the burning gas, just the duration of each pulse.
Would monitoring the gas meter not be more sensible? I believe you can use a reed switch to measure rotations of a magnet within the meter housing.
He must have a really old heater, he could probably save gas by getting a new one installed.
Modern heaters don’t burn gas at a constant rate, instead they regulate the gas flow to keep the water at a constant temperature.
I’d go for ocr on the gas meter, much more accurate.
Yeaaaaah…seems kinda like a bit of overkill there to me, and I am sure if Murphy didn’t say it, someone else did: the more complex you make something, the less accurate it becomes and the more prone to failure it becomes.
I would say either measure it inline before it gets to the heater, or OCR on the meter itself…this seems like a good way to get something that correlates to absolutly nothing
Uhm, yeah I’d vote for monitoring the actual gas meter, simply because the water isn’t always the only gas powered device (furnace, stove, etc.?) unless of course the above devices are electric.
I’m with everyone else on this — definitely overkill.
Ok, as the creator of this a few answers to overkill/accuracy.
Yes reading the meter directly would be more exact but here not posible. This is a flat in a house with 14 of them. They are 5 stories below me in a public room. It wouldn’t be a good idea to put a mobile phone there (it would be away in no time).
Also: Bluetooth wouldn’t work over 5 stories.
And it seems to be pretty accurate. Compared with the value from the meter I am only 1.3% off. (And: even those meters are allowed to have a diff of 2% by law here)
And yes: This heater is from 1991. Getting a new one is no option for me, it belongs my landloard. And I bet he will not buy me a new heater..
And don’t forget: This was fun, find out if it can be done. Nothing what I expect everyone to do.
I think this is way way too overkill!
It could’ve been done with a microcontroller and a special light sensor (maybe with a blue filter?).
Why have a camera with a megapixel resolution when all you need is one ‘blue brightness’ value?
i think is kind of overkill, too.
I think it is really cool, and not overkill, just using the equipment you have lying around. If that is a cameraphone, you use that.