How BlueZoo, Google developed 'building telematics'

di-ny-skyline-buildings-stock-071320

Telematics transformed the auto insurance industry, introducing a new usage-based insurance model and shifting risk measurement from actuarial-based assessments to more precise, data-supported risk quantification. Telematics collects and transmits driver and vehicle data via a GPS tracking system, sends the data to a server that interprets the information and makes it available to end users. Auto insurers can then price policies according to actual usage and identify driving behaviors that may indicate increased risk.

BlueZoo, a Menlo Park, Calif-based startup has developed a building telematics solution that works much the same way, but measures commercial property risk instead of driver behavior. BlueZoo serves commercial general liability carriers, helping them predict risk and proactively prevent claims in office buildings, hotels and other spaces.

BlueZoo has innovated and built on Google Cloud, using its data and artificial intelligence products to more accurately measure building occupancy to help property owners reduce insurance costs, mitigate risk and improve building safety and help insurers more accurately price risk. BlueZoo uses anonymized Wi-Fi signals from mobile phones and optical sensors enabled by Google Cloud.

The BlueZoo offering uses small on-site sensors similar to the dongles outfitted in automobile telematics solutions to monitor and measure building occupancy and determine people’s movement within indoor spaces. The company uses cloud services to transmit and make sense of the data for actionable insight. Carriers can then determine people-related risks by gauging how densely they are packed into spaces, how they’re behaving and moving and if building or space occupancy exceeds fire code guidelines. Buildings with fewer occupants are less risky than those that are much busier and have dense foot traffic.

The technology combines the cloud’s compute power and storage efficiency for training machine learning (ML) models with robust edge inference to deliver precise and timely data, continuously measuring lighting usage, foot traffic, congested areas and other risk metrics with 99 percent accuracy. The optical sensors contain Google’s Tensor Processing Units (TPUs), enabling data analysis at the edge without revealing personally identifiable information to boost performance while maintaining regulatory compliance.

“Google makes available to us technologies that a couple years ago were not available from any supplier,” says Bill Evans, BlueZoo’s CEO. “Google’s combination of cloud technology and AI technology permit BlueZoo to release a new generation of occupancy sensors that do a better job of measuring occupancy than was possible before. If not for the technology foundation that Google was providing to us as a solutions provider, these products would not be possible and if not for Google’s TPUs, we wouldn’t have been able to develop our optical sensors.”

According to Henna Karna, managing director, Global Insurance & Risk Management Solutions, Google Cloud, Google strives to build enablement capability. “This is a great partnership for us because when we see the capability being leveraged for the right use cases and the right mission, that’s exactly what the benefit of Google is,” she says. “It’s leveragable and accessible at scale for a plethora of use cases and that’s where our power comes in. In this case, we have the ability to impact underwriting.”

Typical use cases include not only risk detection and mitigation but premium pricing, new product development and identification of most and least desirable customers. “The opportunity for insurers is to distinguish their least risky customers from their most risky customers,” Evans points out. “The least risky customers are subsidizing the riskiest customers, but they don’t know which are which. If they can distinguish the two groups, they can offer discounts to the least risky and either raise the premium price or collaborate with the policyholder to mitigate the risk for the more risky companies. If a hotel is deemed riskier because it constantly has 50 percent more people in its ballrooms than the fire marshal thinks is appropriate, the insurer can identify the problem and work with the property to reduce the risk.”

This capability will help insurers price better, and also helps them better understand risk. “Where insurers maybe couldn’t write risk before, they’ll know more and be able to come up with definitive ways of looking at risk factors,” Karna adds. “Then carriers can move from understanding risk to actually working with property owners to prevent risk.”

BlueZoo also uses Google Cloud’s Vertext AI, a unified platform for ML operations, BigQuery, its scalable and serverless data warehouse and Looker, its embedded platform for business intelligence and data analytics.

For reprint and licensing requests for this article, click here.
Telematics
MORE FROM DIGITAL INSURANCE