7 Industrial IoT Startups Using AI to Monitor Machines
If you follow technology news and trends long enough, a few statistics pop up again and again. One oft-quoted stat goes something like this: About 90 percent of the data in the world has been created in the last two years. Now, we’ve been hearing this factoid for years, so it’s possibly time for some underemployed MBA to redo the math. However, the bottom line: There is a isht-ton of data out there today, and more and more companies are developing ways to corral and coax that information into valuable insights for a price. Combining these big data sets with artificial intelligence, specifically machine learning, has given birth to an entire industry of predictive analytics. Add the Internet of Things (IoT)—devices from coffeemakers to cars connected together with sensors and beacons—and you can start to detect patterns and make predictions about each network. Like your smart toaster is about to burn the house down.
In the world of the Industrial IoT, there are potentially billions of dollars that could be saved if companies could anticipate machine breakdowns or failures before they happen. As we’ve noted before, General Electric (NYSE:GE) has made some big investments to position itself as the IIoT company. Of course, there are more than a few startups hoping they are agile enough to compete against the likes of GE. We introduced you to one such company, C3 IoT, last year after it took in a $70 million Series D. Now, we’re going to introduce you seven more.
On the Uptake
This year’s IIoT mega-round winner appears to be Uptake Technologies, which closed on a $117 million Series D at the end of last month. That brings total disclosed funding for the Chicago-based startup to about $263 million, for a post-money valuation of $2.3 billion. Not bad considering it was only founded back in 2014. Uptake’s “insight platform” crunches millions of data points from its clients to provide actionable information to improve efficiency, safety and asset performance. In real-world language, that means Uptake analyzes data coming in from machines, from oil rigs and wind turbines to locomotives and tractors, to identify possible maintenance problems or performance issues. The system gets better—learns—over time as it compares its analyses against findings from technicians in the field.
For example, in the first week that Uptake deployed its insight platform at a wind farm operated by MidAmerican Energy Company, a subsidiary of Berkshire Hathaway Energy, it identified signatures that a gearbox main bearing might fail on one of the turbine towers. A few hours of downtime to address the issue cost the company about $5,000 versus up to $250,000 if the turbine had completely crashed. Uptake estimates it can save a large wind farm about $3.3 million per year. Similarly, its algorithms have helped save one rail company $80,000 per year per locomotive in maintenance costs. Other high-profile customers include Caterpillar (NYSE:CAT) and its three million oversized Tonka Toys.
The brains behind the operation are Brad Keywell and Eric Lefkofsky, two co-founders of Groupon (NASDAQ:GRPN). Despite the fears of such luminaries as Elon Musk and Stephen Hawking, who believe AI will destroy us all, Uptake is upbeat about the power of machines to do good. It has even established a philanthropic arm of the company that provides predictive analytics to nonprofits at no cost. It has launched a tool, Student Union, for first-generation college applicants to apply to schools where they would have the best chance to succeed. Another platform called ReRoute tackles human trafficking, which calls to mind our recent piece on how AI will help fight crime.
Uptake is certainly one of the heavyweights in IIoT predictive analytics, but it’s not alone. The bright minds at CB Insights put together this impressive market map.
The categories are a bit fluid—you’ll notice that Uptake and C3 IoT occupy different categories despite sharing space in the same category in an earlier IIoT market map—but you’ll see there are at least six other startups competing against Uptake in the “AI, ML, Predictive Analytics” space. Let’s take a quick look at these other players.
Maana from Heaven
Founded in 2012, Maana out of Silicon Valley is one of the better funded startups in the category at about $43.2 million. It also has some heavy hitters in its corner, including GE, Shell and Chevron. Not surprisingly, its business leans heavily in the oil and gas industry. The insights through its Knowledge Platform aren’t just restricted to predicting when a pump might fail or who shot JR Ewing. In one case, for instance, Maana developed an application that helped its customer to forecast supply and demand for liquid natural gas (LNG) markets. The platform constantly sucks in data—including consumer LNG consumption, market data, consumer confidence, weather predictions, industrial production, and historical shipment data—and spits out real-time pricing for each market.
Out of Sight
We first covered Sight Machines last year in our Internet of Everything feature. Since then, the startup has raised even more money, to bring total funding up to about $30.5 million. A recent case study showed how its IIoT predictive analytics platform can make a big difference. Sight Machines was able to help a client, which manufactured pressure sensors and was experiencing huge amount of scrap waste, to save $500,000 in just three weeks. How? The San Francisco-based startup was able to automate data collection, pull in additional data sources, and provide the information that helped the customer track down the inefficiencies.
Founded way back in 2011, New York-based Augury has raised $26 million for an IIoT solution that relies as heavily on hardware as software. Augury makes a gadget that attaches to commercial equipment such as refrigerators and heaters. The device records vibrations and ultrasonic sound and uploads it to Augury’s cloud service, where algorithms go to work to predict the health of the machine, according to a story in Wired. An app provides ongoing status reports and alerts before something is about to blow.
Update 1/31/19: Augury has raised an additional $25 million in Series C funding to expand its global operation and enhance its industrial analytics capabilities. This brings the company’s total funding to $51 million to date. Read more about the company in this feature article.
Fly Like a Falcon
Founded in 2013, Falkonry is another Silicon Valley startup. The company has picked up $6.3 million in funding. It’s backers include Basis Set Ventures, one of a number of emerging VC funds focused on AI. The company’s machine learning system, Falkonry LRS, detects patterns in data that can help avoid costly downtime. In one case, Falkonry applied its algorithms to help a mineral production company avoid downtime on its processing line due to variations in raw material. One hour offline equals $30,000, so time is money. Falkonry crunched data from motor currents, temperatures, valve settings, and other measurements to provide an early warning system to identify bad raw material conditions that could shut down a processing line.
Steady as a Rock
At first we thought Alluvium was a sequel to the sci-fi blockbuster Elysium. Alluvium is also a term that refers to “a deposit of clay, silt, sand, and gravel left by flowing streams in a river valley or delta”. On the third try we found that it’s yet another IIoT predictive analytics company, this time out of hipsterville Brooklyn. The company, founded in 2015, has raised $2.5 million, with investors including Lux Capital. It calls its platform Primer, which crunches historical operational data to show how stable a factory operation was at any given time. A “Stability Score” calls attention to moments of instability that can be addressed by the tech team to improve operations.
Our entrant from Startup Nation (ie., Israel) is Presenso, which has pulled together $2 million in disclosed funding since it was founded in 2015. The company presently serves four industries with its machine agnostic platform (meaning it learns and adapts to the different sensors and different machines): automotive, oil and gas, energy and water. By now, you’re probably getting the idea of how this works, but here’s the CliffsNotes on how machine learning predicts machine failure.
Eitan Vesely, CEO of Presenso, goes into further technical details about IIoT cost savings (up to 30 percent less downtime), by using the example of a wind turbine here.
Conclusions on Industrial IoT
Sometimes it’s hard to parse the sectors where AI is truly making a difference and where it seems more like fancy window dressing, like picking fashionable pantsuits. But we’re seeing some real-world savings when it comes to IIoT predictive analytics. It’s no wonder that investors poured $2.2 billion into all of IIoT in 2016, according to CB Insights. Companies like Uptake and C3 IoT seem like legit contenders to pull off an eventual IPO, if they don’t decide to make a quicker exit through an acquisition. Some of the smaller funded players, on the other hand, might make an attractive pick up by companies like GE and Cisco who are looking to shore up their dominance in IIoT. It will be interesting to see how the sector shakes out in a few years.
Are you paying too much in transaction fees to your broker? Check out a brokerage firm called Zacks Trade that's offering $1 trades for U.S. stocks and options until 2019. After that, you'll pay just $3 a trade or a penny a share, whichever is greater. It's one of the cheapest brokers out there and you can also trade stocks on foreign stock exchanges. Trade US stocks and options for as low as $1 per order until July 4, 2019.