Whisker Labs, the company behind the Ting sensor that plugs into homes and monitors for electrical faults to prevent fires, has expanded its network of residential sensors to the point where it can now start offering insights to utilities about the performance of the grid.
The company, which sells about 50,000 Ting sensors each month through insurance companies, has managed to build an IoT business that’s sustainable, useful, and completely invisible to most consumers.
Bob Marshall, the CEO of Whisker Labs, told me that the company now has just under 250,000 devices in people’s homes thanks to deals with State Farm, Nationwide, and other insurance companies. Insurers offer Ting devices to policyholders because they help prevent fires from electrical shorts. I personally own a Ting and pay for the $50 annual service myself because I’m terrified some of my DIY wiring will cause a fire.
I spoke to Marshall this week to get an update after having him on the podcast a year ago. In that show, we talked about pivoting the company from weather monitoring to building a sensor and algorithms that could detect electrical arcs and faults, and how he managed to convince insurance companies to buy the device and send it to their customers.
A year later, Whisker ships 50,000 Ting devices every month and has prevented 3,500 fires so far (the service prevents an average of 10 fires a day), according to Marshall. When the Ting device detects a problem, a Ting fire and safety expert calls the homeowner and helps them find the problem. Ting will even send out an electrician to fix an issue. For insurance firms, the data has shown enough savings to make the cost of the Ting devices and the annual service worthwhile.
While Ting has been focused on tweaking its algorithms to provide even more details about potential fire hazards in the last year, it has also started working with utilities to share data about their electrical networks. Every Ting sensor tracks not just electrical variations within the home, but also variations and power issues coming into the home.
This gives it a good sense of the health of the overall grid. And at a time when utilities are trying to manage fluctuations in supply caused by renewables, increases in demand from electric vehicles, and challenges caused by climate change, having more information is helpful. “The grid is not designed for the two-way flows of power,” Marshall said. “It is incapable of managing the future right now.”
Whisker Labs is currently working with an undisclosed utility in California to monitor potential problems with transformer boxes, power lines, and substations. The home network provides a lot of information, but adding more data from the utility will improve the algorithm’s accuracy, said Marshall. Anytime Ting tracks a problem, the company’s data science team pulls all of the data and does a postmortem to understand what the raw data showed and what the actual issue was. In this way, Ting’s algorithms continue to get more accurate and descriptive of a problem over time.
Today on the grid, Marshall estimates the company is at a two or a three out of 10 in terms of the depth of its insights, but every time it gets more info from the utility that number will start to go up. For a look at stats on utilities that Ting offers consumers, check out this map, which Ting updates monthly based on power outages, surges, and brownouts in a given region.
The map reminds me of one that’s produced by Generac, maker of home generators and owner of Ecobee. Its power maps are based on data from its generators coming online in the wake of a power outage, but the idea is similar. However, Whisker Labs’ plans to track the grid for failures before they occur is particularly compelling. Parts of California already see power outages when fire risks are high because utilities don’t want a stray spark to ignite a blaze.
If Whisker Labs could detect those sparks in advance, consumers could keep their lights on while utilities could prevent the sparks that generate fires. That would be putting the IoT to very good use indeed.
Update: The version published in the newsletter on Friday, March 10 stated the incorrect number of Ting devices in homes. It is just under 250,000 not 750,000.