I brush some dead leaves from the top of the slab of cement that covers my curbside water meter and pry it off with a long screwdriver. I’d never really looked inside this underground box before; it’s not as grimy as I thought it would be, though there are a few cobwebs that I clear away. There’s not a lot of room in there, but enough for me to easily strap what looks like a giant fitness monitor, stretchy band and all, around the meter. I nestle a plastic canister, containing batteries and electronics, in the dirt next to it and pull up an antenna almost to ground level. For a moment, I wonder what my meter reader is going to think about all this, but then reason that, living in Silicon Valley, he must be used to odd gadgets. I replace the lid.
From now on, I will be able to know about every drop of water that flows in my house, and exactly in what way many of those drops are being used, by checking my phone. At least, that’s the promise made to me by Marcelo DeCamargo, founder of Nudge Systems, and one of the designers of the device I just tucked into the ground, the Pleco.
From DIY to Startup
Marcelo DeCamargo began his career as an electrical engineer in Brazil, eventually moving into management and to the United States. He kept his hands in engineering as a hobby, and, about seven years ago, decided to automate his home as much as he possibly could.
At the time, Alexa was still a baby and home automation a patchwork industry. So DeCamargo built many of the gadgets he needed from scratch, including circuit boards for his hub and gateways to connect sensors for temperature, humidity, lighting, and motion—including the opening and closing of a pet door.
Then one day in 2014, he told a college friend, Nelson Pedreiro, about his project. Pedreiro’s wife, Ana, works for a utility in California, and with the state in the midst of a drought, ways of encouraging water conservation were top of mind. She asked if DeCamargo’s system could give real-time information about water usage that could help consumers reduce their water use.
DeCamargo couldn’t find anything on the market that did that without requiring an invasive plumbing operation. So he and Pedreiro began experimenting. Instead of measuring the water flow itself, like the invasive systems, they came up with a plan to piggy-back on the measurements being made by a typical mechanical water meter.
“In the existing device,” DeCamargo explained, “there is a part inside that turns with the water flow. That part is magnetically coupled to the dial that shows the reading.”
The two started with an off-the-shelf magnetometer, the kind that’s used in cell phones to help orientation in navigation apps. They added a processor and built the chips into a device that straps tightly around the water meter, like a big Fitbit. A wire from the sensor/processor package connects it to a box containing D-cell batteries that will last for about a year, a radio transmitter, and a directional antenna that, for underground meters, sticks up above the meter itself but remains hidden beneath the meter’s cover. Data is sent on the 433 MHz band to a receiver in the house; the signal will carry at least 50 feet, through walls and concrete, and much more in many cases, DeCamargo says. In about a year, the two had a working prototype. DeCamargo patented the technology, sharing credit with Ana Pedreiro.
Turning the data gathered by the magnetometer into useful information about how water is used in a home took what turned out to be another two years of development time, including figuring out how to filter out background noise and testing tens of water meters to determine how to convert magnetic pulses into gallons for each model.
“It turns out,” DeCamargo said, “that there are only a few different variations. To identify which one a user has, we are shipping our device in a container that holds a gallon of water; we ask the user to fill it as part of the calibration process.”
To develop the algorithms that distinguish, say, a toilet flush from a shower, they built 20 prototypes, and distributed prototypes to friends and family, asking the users to log their water use. Every household showed very different signal patterns, thanks to differences in piping configurations, but, said DeCamargo, they were able to spot common features. Machine learning helped develop the algorithms to distinguish the water use of showers, faucets, toilets, irrigation and a general “other” category. Research on how to decode specific types of water use from the overall flow continues, including ways to identify dishwashers and other appliances, whose water use varies dramatically between models.
A key selling point for a water monitor is not, however, normal water use data, as helpful as that can be for identifying ways to conserve water in a drought. It’s for detecting a leak.
And leak detection, it turns out, is the easy part. “A leak is a continuous baseline flow,” DeCamargo says. “In most homes, water goes on and off, there is no reason to have a constant flow.”
In 2015, the two incorporated Nudge Systems to commercialize their device, which they call Pleco, after the aquarium fish it resembles. The company currently has three full-time employees, relying on contract engineers for much of the ongoing development. DeCamargo and Pedreiro direct the engineering and marketing while continuing their day jobs, DeCamargo as an engineer in the geothermal energy industry and Pedreiro as an executive in the aerospace industry. They are bootstrapping the company without outside funding, filling orders for 1200 devices to date.
“This approach has made us a bit slow,” DeCamargo admits. And, indeed, while the two were taking their idea from concept to product, another company was likewise inspired by California’s water shortages to bring a similar product to market: Flume. Flume, founded in 2016, went the traditional venture capital route, with $7.5 million in funding to date.
Like Nudge Systems’ Pleco, the Flume device straps onto a water meter to track changes in the magnetic field and turns that into digital data sent to a receiver in the home. The Flume2 at US $199 is a little cheaper than Nudge Systems’ Pleco Water Watch at $249. It doesn’t include a screen on its Wi-Fi bridge for at a glance water tracking—honestly, that’s not a feature I found myself using when I tested the Pleco, we’re all so smart-phone bound these days, it was easier for me to use the app. DeCamargo suggests that the second generation Pleco will be designed without the screen, reducing costs.
The Flume gadget is, at the moment, mainly being marketed as a leak detector. It does give live updates of overall water usage, but does not currently distinguish between types of usage. (A friend who owns the device figures that out by watching the monitor in real time as, say, family members shower.) Flume promises such a feature will be rolled out as an update to its second generation device.
Nudge Systems is in talks with companies about using licensing and other agreements to bring the product to a broader market. DeCamargo says that there has been considerable interest, to date, from outside the United States, including companies in Europe, Australia, and Saudi Arabia. Nudge Systems has also begun looking for funding in order to scale up production itself.
At My House: Eight Weeks Later
After I got that Pleco in the ground, I did have a little trouble pairing it with the indoor receiver, mainly because I failed to strictly follow the instructions, which required that they be separated by 15 feet to avoid signal saturation.
The display, smartphone app, and online tools all display minute-by-minute use for the current day in easy-to-understand graphics. They break down water use by category in real time, and aggregate use, again by category, for day, month and year. The web version adds access to any individual day’s record in the past. With just two people in the house in this work-at-home world, I could easily compare the length of my husband’s shower to mine. (His is longer.)
I do find having a leak detection capability reassuring, and that would be the main reason I would buy the device, finding water savings as we enter another drought summer would just be a bonus.
Meanwhile, DeCamargo and his team continue to work on the software, looking at ways to make the information more useful. They realize that turning a DIY project into a successful consumer product is a long shot.
But, says DeCamargo, “We followed our dreams, and we are making something happen with them. We came up with a product that has some limitations, but we are happy to keep working on it. And we are having fun in the process.”
Tekla S. Perry is a senior editor at IEEE Spectrum. Based in Palo Alto, Calif., she's been covering the people, companies, and technology that make Silicon Valley a special place for more than 30 years. An IEEE member, she holds a bachelor's degree in journalism from Michigan State University.