Urbint pulls in $60M for its AI-powered tool to flag the dangers facing utilities

From pipeline ruptures to wildfires caused by faulty equipment, powering the U.S. grid is risky business.

  • Link copied to clipboard

Hidden in the reports of infrastructure inspections and repairs that utilities work on every day are an array of enormous and often hard-to-calculate risks, ranging from power lines that can start wildfires to gas pipelines that can leak and explode. Extreme weather driven by climate change is only making these risks harder to manage — and U.S. utilities are decades behind on the investments needed to shore up their operations and guard against these sometimes unpredictable threats.

Technology that can reveal and rank these risks can give utility workers advance warning of truly dangerous or even deadly accidents waiting to happen, as well as imposing some semblance of order on the endless backlog of work needed to keep modern energy systems running. It could also help ensure that the hundreds of billions of dollars in infrastructure investments being contemplated at the national level are put to the most cost-effective use.

Subscribe to receive Canary's latest news

That’s the rationale that Urbint CEO Corey Capasso offers up for his startup’s artificial-intelligence-informed risk management platform for utilities. On Tuesday, the New York City–based startup raised a $60 million Series C round from investors with a vested interest in solving these kinds of problems, including Energize Ventures, utility American Electric Power, the venture arm of utility National Grid, and utility-backed Energy Impact Partners.

There’s a lot of risk out there right now — problems that have existed for 100 years and problems we’re just now seeing due to climate impacts,” Capasso told Canary Media. The new funding round, which also included participants OGCI Climate Investments, Blue Bear Capital and Salesforce Ventures, brings Urbint’s total raised to $109 million. The company will use the funds to improve and expand the scope of its risk analysis, as well as tackling new sectors such as telecommunications, he said.

Startups step in to help utilities manage risk

Urbint’s utility customers include National Grid, Southern Company, Dominion Energy and Con Edison, which are using its technology to assess and predict risks to their natural-gas pipeline networks. Other utilities are using the platform to analyze the interplay of infrastructure wear and tear and weather conditions that can lead to the type of power line failures capable of sparking devastating wildfires like those that ravaged California in the past few years.

Utilities have been tapping data and software to manage and analyze infrastructure risk for decades, using a mix of well-established asset management systems from vendors including General Electric, Hitachi ABB Power Grids, Siemens, Accenture, IBM, Microsoft and Oracle. Many of these companies are making the most of advances in machine learning to absorb and analyze massive amounts of data, and then using those findings to seek out and separate run-of-the-mill risks from truly catastrophic ones. So are a number of startups that are bringing artificial-intelligence applications originally developed for other industries into the utility space.

Urbint launched in 2017 with plans to target asset risk in buildings, Capasso said. It was the pursuit of that opportunity that led to a chance meeting with New York utility Con Edison, and the cliché lightbulb moment” that revealed how Urbint’s particular focus — combining data sets from the built environment and the natural world — could help detect risks in natural-gas pipelines exposed to a variety of conditions and contaminants.

Urbint’s purpose-built machine-learning algorithms combine masses of historical data on how different infrastructure has failed in the past, and under what conditions, to inform its predictions of the likelihood of future failures. Things like soil conditions and extreme weather events — those can affect assets in real time, but they also have a residual impact,” he said.

To be of most use to utilities, this process of analyzing the past to predict the future must also deliver predictions that utility workers can act on, according to Capasso. The key is to get to a level of granularity where you’re not making generic predictions.” The ideal outcome is being able to say, This asset or this worker at this place is going to create the risk.”

Building risk awareness into the utility ecosystem

To ensure that the resulting predictions actually get communicated to utility workers, Urbint integrates them into daily work orders, inspection schedules and alerting systems, he added. The company also assesses the relative danger of damages or lives lost to each potential risk in order to prioritize which ones must be dealt with immediately.

As an example of how that works, Capasso cited a relatively recent incident in which Urbint warned a utility to send inspectors to a worksite after its analysis flagged the operation as having both a high risk of pipelines being damaged by excavation equipment and a high risk of posing danger to a nearby nursing home.

It’s hard to precisely quantify the value of these kinds of interventions; after all, they’re measured against a hypothetical world in which they didn’t stop accidents from happening. But some of Urbint’s utility partners have been providing feedback. National Grid estimated that it reduced gas pipeline damages by 22 percent from 2019 to 2020 by using Urbint to prioritize its management of the work-order tickets issued to monitor excavations for pipeline risks.

Beyond this use case of damage prevention, Urbint offers asset integrity” services to identify the most leak-prone pipelines in a natural-gas network. It estimates that its utility customers using this service have been able to prevent leaks of methane, the primary ingredient of natural gas, in amounts equating to more than 60,000 metric tons of carbon dioxide last year — an important metric given that methane leakage is a significant driver of climate change.

More recently, Urbint has been working on identifying power line failure risks that can cause wildfires. Our technology is used to help identify those worn assets that have not been inspected or maintained and are at high risk of failure,” he said. While he wouldn’t name which utilities were using the company’s system for this purpose, he did mention that one use is informing utilities how to manage planned shutoffs and fire-prevention grid outages, strategies that have been in frequent use by California utilities in the past three years.

Fine-tuning funding allocation with smarter risk assessment

U.S. utilities collectively spend tens of billions of dollars per year on natural gas and electricity distribution and transmission infrastructure. That spending will have to increase dramatically in the coming years to support the country’s push to electrify transportation and heating, to connect far-off wind and solar power resources via transmission grids, and to increase resiliency against heat waves, cold snaps, hurricanes, floods, droughts and other extreme weather events intensified by climate change.

Federal action is likely to increase the amount of investment already being ordered in state-by-state regulatory mandates for utilities to make their electric and gas grids safer and more resilient. The infrastructure bill passed by the U.S. Senate last month contains tens of billions of dollars in funding to bolster power grid resilience, and the reconciliation package being promoted by the Biden administration and Democrats in Congress contains hundreds of billions of dollars.

These infrastructure companies are in very tough positions, and they’re trying to do the best they can,” Capasso said. It’s in the best interest of society to figure out how we do our part to engage in this transition and get to a better place — especially as we see these extreme weather events occur at a higher frequency.” Better data collection and reporting will go along with these hefty but much-needed investments, he predicted. It can no longer be ignored.”

(Lead image courtesy of Urbint) 

Jeff St. John is the editor-in-chief of Canary Media. He covers the technology, economic and regulatory issues influencing the global transition to low-carbon energy. He served as managing editor and senior grid edge editor of Greentech Media.