In the absence of specific information about individual risk, insurers have traditionally had to underwrite risks by category. The Internet of Things (IoT) is changing that because of its ability to monitor insured property or track policyholder behavior. This more specific or personalized way of looking at risk is growing in popularity, allowing policyholders to pay for insurance based on the true nature of the insured risk, instead of being charged extra for other people’s bad behavior. As technology and connectivity advance, insurers in both the consumer and industrial markets are finding novel ways of identifying and pricing risk.

Usage-Based Insurance (UBI)

Smartphones aren’t just for texting and looking up directions anymore. Auto insurers can now use them to track your driving habits to better assess the risk and potentially save you money on your insurance. Some companies are also using small IoT-connected boxes, called dongles, that attach to vehicles to track mileage, speed, braking and other habits to determine risk. Now, no matter how you drive, you can pay for just the risks you take.

UBI is also changing the face of the industrial sector. Connected devices attached to machinery are creating methods to track exactly what occurs in a facility or plant. Insurance companies are looking at usage patterns and safety measures collected by on-site sensors that relay critical information around risk. For example, certain plants have earthquake detectors that can automatically shut down powerful machinery to prevent damage or accidents during an event. Likewise, smoke detectors and other ambient sensors can track temperature, toxins, mold and more to combat risk and minimize exposure or damage.

UBI and PdM
Insurers have long relied on aggregate historical data to predict accidents, malfunctions and outages. But relying on “in general” and “on average” costs insurers money. Instead of insuring for what could happen, large policies can now be based on what industrial facilities are actually doing to take care of all their resources, thanks to the new generation of predictive maintenance (PdM).

PdM is an ongoing maintenance approach that does away with the “run to failure” mentality that costs both insurers and manufacturers money. With sensors that track vibrations and ultrasonic emissions from machines, PdM technologies can predict when parts will malfunction or fail far more accurately than periodic maintenance can. This enables companies to intervene at the right place and time to avoid catastrophe. Additionally, parts can be ordered when needed instead of having to keep an expensive stock of spare parts on hand or worse, having a critical machine go down and having to track down a much-needed part. PdM happens in real time, letting technicians know when to repair or replace parts. This cuts downtime and labor costs while streamlining efficient energy usage with fully functioning equipment.

Since PdM can lower the risk of mechanical breakdown by 75 percent, insurance companies can remain competitive with reduced premiums, which in turn leads to happier customers. Combined with approaches such as Equipment-as-a-Service (EaaS), where equipment providers are freed up to focus solely on upkeep and uptime, insurance risks plummet.

PdM as the Insurer’s Holy Grail

The IoT is one of the biggest boons insurance actuaries have seen. The sensors a facility uses to track equipment health and service needs in real-time also serve to assess risk—essentially the holy grail of insurance. The factors used by insurers to determine and set rates can reflect the exact practice a facility uses to manage workflow, machine health and safety, which leads to less downtime, decreased risk and lower insurance premiums.

When calculating risk revolves around continuous diagnostics rather than scheduled preventive actions, the pricing of risk becomes far more precise. Insurance becomes an incentive to modernize in the industrial sector. In the end, it is about efficiency and increasing profits.

IoT Will Force P&C Insurers to Change Business Models—Celent Study