DataCenterNews US - Specialist news for cloud & data center decision-makers
United States
From the driver's seat to the control room: what motor telematics can teach power generation insurance

From the driver's seat to the control room: what motor telematics can teach power generation insurance

Thu, 19th Nov 2026 (Today)
Alistair Moodie and Jonathon Valentine
ALISTAIR MOODIE AND JONATHON VALENTINE Chief Product Officer and Global Chief Technology Officer SAMP Risk

Jonathon Valentine, Global Chief Technology Officer at SAMP Risk, spent his career in motor insurance, starting as the first employee at Insurethebox, one of the U.K.'s first telematics-only insurers, and later co-founding ThingCo. Alistair Moodie, Chief Product Officer at SAMP Risk, is a mechanical engineer who spent more than a decade on power plants working to improve their reliability. Following their recent CII webinar, they explain how the lessons of motor telematics are reshaping how power generation risk is understood, priced and managed.

Whether it's a car on a driveway or a wind farm covering square miles of land, there is still a person in making decisions about how the asset is operated and maintained. Telematics changed motor insurance by giving insurance providers a view of those decisions. The same approach is now doing that for power generation, at a time when renewables and data centres are creating far more sites than the insurance market can survey in the traditional way.

What telematics changed in motor

Early telematics products in motor struggled because they were designed around what insurance providers wanted to sell, not what customers wanted to buy. What made the difference was building every process around the data, so it could be used to its full extent rather than managed as a niche. That data removed proxies. Drivers routinely misjudge their mileage; with a box in the car, an insurance provider can price for the miles actually driven. It made it possible to reward good drivers, nudge risky ones through automated messages, establish fault quickly after a crash, spot fraud and make better decisions at renewal.

How power generation risk is assessed today

Power generation insurance, by contrast, still relies on periodic snapshots of how sites are run. Most power plants and renewable sites are surveyed once a year at most, and often only every three years. We recently saw a site seeking cover on the basis of a six-year-old survey. With renewables and data centres adding sites far faster than the number of surveyors is growing, that gap is widening.

The survey report, typically 40 to 100 pages, travels through the broker to the underwriter, who may not have time to read it in detail unless there is a claim. It records loss estimates, risk recommendations and claims history. However, claims history is not the same as what happens on site.  On a power generation site there is always something going on and very little of it reaches the insurance market. Underwriters are left to price on proxies such as region, technology and age, with conservative assumptions built in. In motor terms, that is like pricing every driver the same way because nobody has fitted a box.

Treating a power plant like a connected car

A black box records speed, acceleration and braking, using this data to understand how the vehicle is being driven. The equivalent on a power plant is engine-style operating data: how hard equipment is being run and how quickly conditions change. At SAMP Risk, we combine that with inspection data.  We look for anomalies, including slow ones. We have seen a bearing failure that developed over three months through a change in the data too small for anyone to spot without the right analysis. Machine learning models learn each machine's normal behaviour and flag when it drifts. Statistical models then translate each signal into a probability of failure for that specific plant, and the results are combined into a simple red, amber or green rating or score.

The differences can be stark. In one example, two identical steam turbines differed only in how well they were looked after: the turbine whose lubricating oil was not being tested was around 12 times more likely to fail. In another, a transformer running hot was around seven times more likely to fail than an identical unit at normal temperature, because heat rapidly degrades its insulation.

Rewarding sites that manage risk well

Identifying risk only helps if someone acts on it. When a site sees that a component is pushing its risk up, they can fix it and see the score improve. Just as importantly, so can the broker and underwriter. At one site, the risk algorithms in our platform flagged two transformers close to failure. The site team fixed them, avoiding what would likely have been a multimillion-pound claim.

Telematics-based insurance for power generation sites also enables incentives to be offered for a site to improve its risk and cut the price it pays for insurance. Motor telematics showed that those incentives need to run continuously: drivers who earned a discount after a few months of monitoring soon slipped back into old habits, just as power sites tidy up before a survey and let standards slide until the next one. A share of the premium is set aside, and the site earns it back month by month according to its risk score. Rather than load the premium for uncertainty about a site underwriters can't yet see clearly, they can turn that loading into an incentive the site earns back by proving it is well run.

Freeing up capital

Better risk insight can also enable new products. One concentrated solar power plant had a 90-day waiting period on business interruption cover for machinery breakdown. Its lenders required it to lock away 90 days of revenue in a reserve account against that risk. Using the plant's data, we helped enable a deductible buy-down that halved the waiting period to 45 days, releasing cash the owners could reinvest.

One year on, the results at that site are encouraging: a 14% reduction in the modelled probability of a major failure, and a 61-percentage-point improvement in how thoroughly the site investigates and records events. We have found that a power plant that carefully investigates the small things is more likely to act promptly on the big ones.

Getting the data, and sharing it safely

Every telematics market has had to settle who owns the data and how it can be used, and clear policy wording is essential. Security matters just as much when connecting to power infrastructure. Our telematics device is designed security-first, independently penetration tested and built to meet the security requirements that regulators in the U.K. and U.S. set for power plants.

We also standardise data at source, so every site's data arrives in the same format whatever the equipment. Owners are sometimes wary of underwriters reading too much into raw data, so underwriters see aggregated risk scores and ratings rather than every reading.

A shared benefit

In motor, drivers who slowed down also saved fuel. In power generation, a plant that breaks down less generates more power and more revenue, on top of better insurance terms. The interests of plant owners and insurance providers are closely aligned.

For insurance providers, the key is to use the telematics data from power generation sites to understand risk more accurately and offer the benefits to policyholders. From there, the market can build new products such as deductible buy-downs, and in time use the same data to make claims faster and smoother.

Annual surveys and generic proxies served the market when sites were few. As distributed generation grows, continuous, data-driven insight is how the market will keep pace.