I am nearing completion of an on-and-off project for the last few months. Using cheap Internet of Things (IoT) components, such as flow meters and Arduino boards to compile a universal meter automation solution. The theory is that you can take this and install it in line with your existing (sub)meter and be able to track your water usage in real time over WiFi or Cellular connections, depending on whatever spectrum(s) you have available at your park.
My goal is to rely upon the data for billing purposes, and true up with the meter readings once or twice per year manually. This solution is setup now to upload data to my Google Drive spreadsheet every 15 minutes so I can access it anywhere in the world on-demand. Each meter (MHP pad) gets its own column, and can add sheets to the workbook for additional locations.
Approximate cost of hardware is about $65.00 per meter, closer to $50 when you do 20+. You might be able to also interface with existing pulse meters and save $10, and will be testing that after this one is completed.
Other functionality: 1) submeter notification (email to our Park’s general mailbox) when continuous usage for 4 hours (water leak) so the manager can chase it down; 2) submeter notification when there have been no readings for a week (either technology issue or resident has moved out); 3) notification on a specific date when the monthly meter readings have been compiled and a new spreadsheet for the new month created.
Reliability (and overhead due to unreliability) of a solution like this is the major concern and I have a weatherproof enclosure bought from Home Depot or Amazon to hopefully address it (more in the next 30-60 days). Also ease of setup for non-technical people will be a challenge in some cases.
Beyond that - is there interest from the broader community in something like this? Are there other key items that I am missing? Does this seem attractive compared to some of you with the proprietary systems today?
I worked on an andrino based meter solution a couple of years ago. Application was slightly different. I needed to have a daily total flow reading on a wastewater treatment plant. Meter out put was 4 to 20milliamp. The software side of the application worked flawlessly but I had serious reliability issues on the hardware side converting the 4 to 20 to milli volts for the board to read. Scraped the project but I still think the concept was solid. I needed a daily flow reading for EPA compliance so since I could crack the reliability nut the project died. I ended up buying on off the self 4 to 20 data logger.
Yea saw similar issues with a couple of the meters on my end when I first started too due to the low voltage, which is why I have looked closer at an in line “Hall effect” flow sensor - these are powered by the board and don’t have the issues these meters have. The key is to make sure that the calculations by the flow sensor and the water meter are the same - or close enough to meet your needs.
This one here has 477 pulses per Liter, which I have the Arduino code convert to gallons and feed to the Cloud. This is what helps make it “universal” and workaround all of the different specs of the 100+ meter manufacturers.
If it works out happy to share the design and code for those to give it a go…
Regarding the “existing pulse meters” that is of interest to me… however these (mine: WM-NLC lead free brass water meters & WM-PC plastic water meters) are a much lower frequency… only one pulse per gallon. It is a simple dry contact reed switch that is closed by a magnet on one dial - so it may remain closed for a long time for a single pulse. (reed switch pulse output details) Will your system accept that as-is (referring to hardware)? if so, you could make that an option/setting in the software. Actually that would be a great feature to make it truly UNIVERSAL. Since that is how ALL of my meters work (once I add the switch), I would certainly be interested in trying your system if and when it is ready.
I would also be more comfortable with that since it uses the actual meter. Therefore: No problem regarding reading discrepancies, reduced hardware costs, and extremely simple installation (just screw the reed switch to the meter).
Let me know when the pulse meter version is ready… Thanks!
Yep, will let you know how this goes once we get the first version working. I have been testing with certain resistors and interrupts that can convert the pulse into a readable dialect for the board, but not with confidence yet.
Right now I am just getting the flow sensors in place that can take the meter-specific variables out of the equation. Below is basically what this looks like before getting into a proper case and installed into the ground.
This is not my specialty at all, but isn’t one pulse per gallon much too low for a residential meter? Like, the error will be huge and the precision poor?
not at all… do you get billed for less than one gallon? most municipalities bill by the 100 gallons or even 1000 gallons. then whatever they miss in one month, gets made up for the next month since it was ignored for the reading, but still had already been measured on the meter. This “low frequency” pulse output is actually quite common in residential and commercial water metering. Keep in mind, that there are all sorts of low frequency pulse rates of this type. 1 per gallon, 10 per gallon, 1 per 10 gallons, it all depends on what needle or dial that the magnet is on and the switch is mounted over. this page explains it well and has a few examples of these meter types: https://assuredautomation.com/pulse-output-water-meters/ then, on this page you can see the options that are available for each size: https://assuredautomation.com/WM-PC/index.php?p=buy
@jhutson - that’s pretty cool! Neptune said they have a new cellular solution coming out that is a little box (cigarette box size or maybe shoebox) with a 10 year battery - they call it a cellular MIU. They want - $600 per door.
