Canadian Underwriter

Things Connected

October 1, 2014   by Craig Harris, Freelance Writer

Print this page Share

The frenetic pace of technological change often begs the question of the “next big thing” in IT: Is it merely a futuristic vision or imminent reality? In the case of the Internet of Things (IoT), the answer is both.

There are already plentiful examples of IoT in action, defined in a recent report by research firm, Celent, as the integration of three components: physical objects with networked sensors; transmission of information to data stores; and analytic engines that examine the data and provide a feedback loop. Industries ranging from agriculture to automotive to appliance manufacturing have installed devices that continuously monitor the functioning of equipment and send data to a source for analysis and action.

However, the insurance industry tends to rely on actuarial data and static information gathered at the policy-application stage for risk profiling, underwriting, pricing and claims management. When it comes to IoT and capitalizing on the real-time data created by sensors and networks, there is a sense that insurers are still in the infancy of evolution.

“Only a small minority of insurers are investing the resources to do the appropriate due diligence on IoT with respect to the underwriting,” says Kevin Kalinich, global practice leader, network risk/cyber insurance for Aon Risk Solutions. “The majority of insurers base their rates on actuarial data. But you will never have 20 years of actuarial data with IoT because it changes dramatically and in real time,” Kalinich points out.

The property and casualty industry insures a lot of “things.” Gartner estimates that IoT will include 26 billion units installed by 2020, and by that time, IoT product and service suppliers will generate revenue exceeding US$300 billion. Research from Cisco notes that more than 50 billion devices will be connected to the Internet by 2020. Whatever the number, all of those devices will be generating various forms of data – numerical, text, audio, video. One of the key questions: Will any of this help insurers in underwriting or pricing risk?


“For insurance, the parts of the puzzle are just slowly falling into place,” says Donald Light, an insurance analyst and author of the Celent study, The Internet of Things and Property/Casualty Insurance: Can an Old Industry Learn New Tricks? “There have to be sensors, there has to be a network, it has to be secure. Ten years ago, nobody would have dreamed of this. Now, people are talking about the collection and analysis of this data,” Light says.

In an increasingly gadget-oriented society, where personal and commercial lines consumers demand mobility, remote access and interaction between devices, the connectivity push will only continue. This trend will be compounded by the relatively new capacity to monitor virtually anything – checking a person’s heartbeat or temperature with wearable technology, unlocking doors with mobile devices, tracking (or controlling) car movement with sensors, remotely monitoring complex building systems for heating, ventilation and air conditioning (HVAC), security, and smoke detection/fire.

“One opportunity is how insurance is going to leverage data for more precise, customized, one-to-one underwriting,” notes Denise Garth, a partner and chief digital officer for technology research firm, Strategy Meets Action (SMA).

“The bigger issue is how insurers can figure out how to get in the game of IoT,” Gartner suggests.

The space-age predictions of the connected home, driverless car, smart cities, self-monitoring commercial buildings or machinery suddenly do not seem that far off. The IoT opens up a world of opportunities, and a potentially disruptive environment, for the insurance industry.

“If insurers can access this big data information, they would really be able to understand products and pricing around it,” observes Michael Petersen, managing director and national leader, communications, media and technology for Marsh Canada. “I think we all understand there is a great opportunity here – if you get it right, there can be a huge payoff.”

A tangible example of IoT in the insurance world is vehicle telematics. Towers Watson reports the pace of adoption of usage-based insurance (UBI) telematics has accelerated exponentially in Canada in the last year. A survey released in April by the consulting company showed that 56% of polled Canadians expressed a “strong interest” in buying a UBI policy.

“The IoT example that is the furthest along in the insurance industry is telematics in vehicles, in both personal auto and commercial fleets,” notes Light, who in his report divides “things” into four categories – living, moving, stationary and transmitting. “In fact, there is some evidence that commercial vehicles may be better candidates for telematics adoption,” he suggests.

Vehicle sensor devices have vast potential in providing services other than just insurance, some sources say.

“Telematics will expand beyond… just car insurance,” predicts Randy Carroll, chief executive officer of the Insurance Brokers Association of Ontario (IBAO).

“A telematics device that is used for insurance purposes could also be used as an anti-fraud measure or theft deterrent, a roadside assist device or vehicle maintenance diagnostic tool. Vehicle tracking, vehicle recovery and accident reporting could all be enhanced if the proper telematics device is used,” Carroll says.

Garth asserts that telematics and UBI are only scratching the surface of what this technology can do, in terms of business assumptions and revenue models.

“Interestingly, UBI is no longer limited to vehicles, but is now being considered for home, medical and life insurance products,” notes a research brief Garth authored for SMA, The Internet of Things: Creating a Connected World. “Sensors like telematics are all about the connected car, the connected home, and the connected life. UBI is the precursor to a broader impact of sensors and the Internet of Things that will allow us to connect the dots between the data for new customer products, services, outcomes and experiences,” the study adds.


The “connected home” has also seen a spate of recent activity, with some suggesting this is an area ripe with opportunity for personal property insurers.

In June, American Family Insurance and Microsoft launched a “business accelerator” for tech start-up firms focused on home automation. American Family Insurance, the eighth largest homeowners’ insurer in the United States, will provide industry experience, consumer insights and homeowner knowledge to companies focused on building safer and smarter homes, notes a press release from the company.

Last January, Google purchased Nest Labs, a maker of “smart” thermostats and smoke alarms, for a reported US$3.2 billion. Analysts noted the acquisition – the second biggest by Google since buying Motorola Mobility – marked the Internet firm’s foray into the next generation of smart home devices, including sensors and networks for remotely controlled appliances, door locks and other everyday objects.

Lost in the shuffle of the recent iPhone 6 release and the wearable technology of the iWatch, Apple also unveiled HomeKit, its iOS-based protocol for hooking up connected gadgets in the home.

Other software companies have jumped on board, including August Smart Lock, an electronic lock that provides keyless entry into the home through Apple mobile devices.

What does this mean for personal property insurers? Light notes that the information coming from the next generation of Internet-connected devices will provide a slew of data to insurance companies. Is the home well-maintained? Is it secure? Is there a consistent temperature level? Are there moisture detection devices in basements and around washing machines and dishwashers? Are there advanced smoke and carbon dioxide detectors?

“The real question is when insurance companies will have enough confidence in thei
r analysis of the data generated for underwriting and pricing,” Light suggests. “I think IoT will slowly revolutionize how risks are priced and underwritten for property insurance,” he comments.

“The implications are quite interesting when it comes to information and monitoring and what that could mean for loss and risk,” Petersen says. “For example, something like a sewer back-up system could be set up to notify someone on a mobile device, but who is going to be notified? Who will ensure it is turned on and functioning properly?” he asks.

“There may be a benefit, but that data has to be useable; it has to be collected back and used by the insurance company to tailor products that fit consumer needs,” Peterson points out.


The potential of IoT may be even greater in a commercial lines for property, machinery and equipment. Many larger buildings already have sophisticated operational systems in place that monitor HVAC, security, fire/smoke detection and sprinkler systems. In most cases, that data is already being collected and examined, at least by the owners or building management companies.

Light explains that there is no technical reason that information could not be made available, “either directly or in meta form,” to insurers.

“Insurance companies don’t necessarily want to hire an HVAC expert to parse the data for all the clients they insure,” he says. “However, if you have data aggregators outside of the insurance industry, they could, in effect, create safety scores or loss control scores that provide valuable data. This could factor into pricing, risk analysis and underwriting.”

The examples of industries and commercial sectors to which IoT could apply are virtually endless. Aon Risk Solutions’ Kevin Kalinich cites the case of mining in Canada and the real possibility of using robotics and driverless trucks on remote sites in northern regions.

“This is already being discussed in the industry,” Kalinich reports. “I think where the IoT will really change things is in the industrial Internet. Huge machinery, huge operations will benefit from this,” he predicts.

Loss prevention and risk management represent additional areas of potential advantage for commercial clients. “By definition, you have hard-headed business people who may be risk managers or have a financial interest in lowering risk costs,” says Light.

“The IoT value proposition for buildings, operations, general liability, security – these are all areas that companies are very receptive to in terms of improved loss control,” he adds.


Several sources contend, however, that the prime opportunity for insurers in IoT is in information not objects. And that data must be tied to risks, hazards or exposures.

“The big money is less in the devices and more in the acquisition and processing of data,” Light notes.

“The devices in most cases are pretty cheap already. However, putting the data in the proper form for obtaining insight is where the real value is going to be. You have to be measuring things that have a pretty clear relationship to loss or hazard,” he explains.

One of the key questions is how far along insurance companies are in incorporating IoT and big data into their underwriting, rating, risk analysis and claims management/loss prevention systems.

“Insurance companies will have to wake up; they cannot focus only on actuarial data,” Kalinich argues. “There has to be a combination of actuarial experts with technology, privacy and legal experts to understand these risks. The impact of this is going to be felt everywhere – in engineering, design, architecture, commercial devices,” he says.

“It costs an insurance company an awful lot of money to get it wrong,” Petersen comments. “You can spend a lot of money on the software and hardware, but the data could be useless or doesn’t help the product.”

Garth points out that much of the data coming from IoT is from other industries – automotive, manufacturing and industrial. It is also hitting insurers with the velocity, volume and variability of big data.

“I think insurers are very much beginning to understand the importance of big data from the Internet of Things,” Garth says. “But, frankly, there is still a lot of work to do in terms of their systems, common data models and definitions internally. They still have to achieve a level of data mastery,” she adds.

“Mastering” the big data emanating from the IoT is not only an opportunity and challenge for insurers, but also for brokers.

“Brokers embracing data will see a positive return on their investment,” Carroll expects. “The more brokers know about their customer, the easier it will be to meet the ever-changing expectations of today’s consumer. The challenge is understanding what data can do and the risk is not keeping up,” he adds.


To get the data in the first place, insurers and brokers will need to encourage the use of interconnected devices in the homes, cars and businesses of customers for insurance purposes. Much like earlier experiments with vehicle telematics, there is a need to establish a comfort level with clients, as well as address potential issues of invasiveness and privacy.

“Insurance companies will want to know this information, not just about commercial, but personal clients,” notes Kalinich. “IoT can tell who has a carbon monoxide detector in their homes, but it can also track what kind of food is in your refrigerator. These can be positives, but they can also be invasive,” he cautions.

“People need incentives,” Light suggests. “You have those who are gadget-oriented and get excited about new connected devices. Others are going to say, ‘I am paying X dollars per year for my insurance, and I could get 20% discount if I install some of these devices.’ That might be worth it for some people to be early-adopters,” he notes.

The quid pro quo for discounts is data. And some brokers are concerned about who owns that data and how it will be used in the new world of IoT.

“You have to balance premium discounts with the fact that the company owns your data, your profile,” says Kat Macaulay, a digital communications specialist with Rogers Insurance. “What will happen with that data? Could it be sold? That is what consumers need to be aware of when we have this data gathering. Who owns the data, and will my privacy be protected?”

Garth suggests that IoT may usher in a new era of data ownership and stewardship guided by consumers. “There is a view that a lot of this data is going to be owned and managed in a digital environment by the customer, and the customer will authorize access and use of that data,” she says. “They will expect something in return, whether it is a discount or service.”

That represents a dramatically different equation for insurers in the status quo of data collection.

“We are going to have to rethink the purchasing of all these data sources from third-party providers that have aggregated lots of information,” Garth suggests. “We may actually have to consider how we will have to get real-time data from our customers, not just at the policy-application stage.”


Another major obstacle for the widespread adoption of IoT in insurance is the security issue of sharing information across multiple devices and networks. A Hewlett-Packard study released in July found that 70% of the “most commonly used IoT devices contain serious vulnerabilities.”

The survey did a scan of 10 of the most popular IoT devices, such as televisions, webcams, thermostats, remote power outlets, sprinklers, door locks, home alarms, scales and garage openers. On average, 25 vulnerabilities were found by researchers – with a total of 250 security concerns discovered.

“There is a great quote about three simple laws – ‘law one, everything on the Internet can be hacked. Law two, everything is being connected to the Internet. Law three, everything else fo
llows from the first two laws,'” says Petersen. “The interconnectivity that is going to take place presents a very interesting situation. It’s the whole: ‘Do you know who your fridge is talking to?'” he says.

“The fact is that any devices can fail, and any networked process can be hacked,” Light comments. “These risks are probably more acute on the commercial side because the systems, sensors and data they already have were not designed to be secure; they were not designed to be shared. Today, we live in a different world,” he suggests.


For Garth, the greater risk may be that insurers are missing the boat on IoT or merely using the big data from interconnected devices to tweak existing insurance products.

“If, in fact, the devices can be attached to things in the home, car, business, we can begin to create new services and product offerings that actually provide real value,” Garth observes. “They can help you reduce your risk and potentially even eliminate your risk. Those services can become a driver of the relationship for insurance, so the product becomes more of an add-on. I think it is an opportunity to re-imagine the product and the business model for insurance.”

The potential threat from other industries or sectors that have widely embraced IoT is whether these companies will be competitors or collaborators with insurers.

“In this new world, businesses are going to begin to offer solutions that have the IoT connected and services tied to that,” says Garth. “I think it opens up an opportunity for other entities to manage the customer relationship, the customer loyalty, the customer experience. Insurance may be just a piece of the bigger picture, a spoke on a hub.”

Garth notes that surveys done by SMA show insurance companies are dipping their toes into IoT waters. In a recent study of emerging technologies, the firm found 20% of insurers surveyed were currently piloting, testing or deploying IoT projects. That number rose to 50% over a three-year future timeframe.

“I put the tipping point at 30%,” notes Garth. “I think there are a lot of insurers doing stuff quietly. But when you have one-third of the market doing something, it becomes very significant.”

(Major p&c insurers in Canada contacted for this article did not provide comment at press time).

Others argue that the evolution of IoT within the p&c insurance environment will be a slower process. How long it takes for sensor-equipped devices to be widely installed in homes, cars and businesses, and whether the ensuing data will be useful for underwriting, pricing and claims/loss control, remain unanswered questions.

“I think the analysis of data for pricing and risk outside of telematics in vehicles is really at an early stage,” Light says. “It is kind of like, ‘Oh, we can get this data, what do we think this could mean in terms of losses?’ There are some experiments and early initiatives going on at some insurance companies,” he reports.

“The reality is there are a lot of things that will have to happen before this takes place,” Petersen points out.

“It has to involve new economy firms working with old economy firms. The insurance industry doesn’t have the money to do it itself; they may be able to buy access to big data to understand the information. They can then decide if they can use it for things like finding the better customers and building better risk profiles,” he suggests.

“I believe we are at the infancy of this, but you can’t just throw up your hands and say, ‘We don’t have all the answers, so there is nothing we can do,'” Kalinich concludes. “You have to think about it. The insurance industry will have to figure out the implications.”