When you think of autonomy, self-driving cars may come to mind. But that’s just one of the many autonomous uses today. Autonomy is constantly evolving, and it has the potential to create changes in insurance and various industries.
“Autonomy is a very broad topic and it can be used in so many different applications,” said Andrew Zarkowsky, global technology industry practice lead at The Hartford. “At first, when I thought about autonomy, I thought about a car driving itself in Boston. But now, I think about the different uses across the country, whether it’s in manufacturing, agriculture or in ocean marine.”
Advances in autonomy can bring increased efficiency, productivity and worker safety. But its full impact on the insurance industry and other sectors remains to be seen.
Because of this, The Hartford created the “Autonomous Technology Working Group” to examine autonomy and how it affects different industries.
“It’s a cross-functional team and includes employees from claims, emerging risks, Risk Engineering and underwriting,” Zarkowsky explained. “We’re trying to hit this from a variety of different angles so we fully understand how autonomy will affect all aspects of the insurance industry. We’re studying risk analysis, our underwriting process and claims. We’re also looking into our coverage and risk management services to make sure they provide the best level of service now and as the autonomous industry evolves.”
What Is Autonomy?
Autonomy refers to operations or actions that can be done without the control of a human. Ideally, autonomous functions don’t require a person to intervene for a correction. To do these functions without the control of a human, it requires artificial intelligence (AI).
“The concept of AI is for computers to think and make decisions similar to humans,” Zarkowsky explained. “LIDAR uses lasers to mimic a human’s eyes and vision.”
How Is Autonomy Used?
One of the most publicized use cases of autonomy that’s being explored right now is within the automotive industry. Car manufacturers – most notably, Tesla – are developing ways for vehicles to control some driving functions with little intervention from the driver.
However, there are many other use cases happening now, such as in agriculture. Certain farm equipment can operate without a driver, which can increase worker safety and improve efficiency.1 When it comes to farm operations, autonomy can make it easier for scaling and growing the business.
Manufacturing companies are also investing in autonomy to become more efficient. One source estimates that the warehouse automation market will reach $41 billion by 2027.2 One factor driving the increase is the rise in e-commerce and the need to fulfill customer orders quickly.3
“There’s a lot of learning happening right now with autonomy,” said Nadia Gluch, the director of The Hartford’s Emerging Insurance Risk Unit. “We have applications in autonomous vehicles, forklifts in the factory, farming or for limited use in truck fleets. Companies are learning how to make a car drive autonomously or how to move packages in a warehouse. They’re taking those lessons and trying to apply them to what can be done with more complex operations.”
What Does Autonomy Mean for Insurance?
As autonomy can improve efficiencies in other industries, it can do the same with insurance. For example, with the added sensors and technology in cars, it can streamline a claims adjuster’s job.
“There’s essentially a black box in cars now. You’ve got all of these sensors, all of these cameras and all this knowledge,” Zarkowsky noted. “We’re entering a world where, from a claims perspective, it won’t take a lot of time to figure out what happened to a car after an accident. We can look at the cameras on the car or a camera in the street intersection. So in some degree, it’ll be easier to figure out these kinds of claims.”
Ryan McKain, a senior consultant in The Hartford’s Claims and Operations Department, added that the technology in cars can provide valuable information during an accident.
“I think we’ll get to a point where reconstructing an accident is going to be easier and more straightforward,” McKain explained.
Autonomy Creates Efficiencies, But Can Also Cause More Complexities
Autonomy and Data
While autonomy can make certain aspects of insurance more efficient and easier to understand, it can also make things more complicated.
An autonomous system is only as good as its data, according to Gluch.
“Something that often gets overlooked is that so much of this technology depends on data and network support. So, how much of this is going to rely on bandwidth and network infrastructure?” Gluch asked. “If the navigation system in a car can fail because of a lost signal, imagine a car losing signal or an entire fleet of vehicles losing signal. All of this demand can put tremendous strain on our networks.”
Zarkowsky added that using autonomy in any financial-based decision-making process, including insurance underwriting, has clear benefits. For example, autonomy can be used for simpler underwriting requests – giving underwriters the capacity to work on more complex projects. Zarkowsky cautioned, however, that these kinds of use cases need to be set up correctly.
“If an underwriter makes a mistake and gets something wrong, you get it wrong once and it affects one person or business. But if you get it wrong with an autonomous program, that AI algorithm can end up impacting hundreds or thousands of people at a time,” he explained. “That severity of risk is why these kinds of systems have to be set up correctly.”
Autonomy and the Question of Liability
One of the biggest questions that autonomy creates for insurance is who’s liable. If someone were using a car manufacturer’s software to help with driving and the driver gets into an accident, who exactly is liable?
“We’re in a spot now where we’ve got a pretty well-defined set of laws and standards and norms. Did you go through the red light? If so, then you’re liable,” McKain said. “But we have to switch that thinking now. Did the accident happen because the radar sensor in the bumper failed or the module that interprets that sensor didn’t work properly and failed to engage the brakes?”
A Shift to Product Liability
In terms of insurance, the rise of autonomy brought on a shift of “negligence from an individual perspective to negligence from a products liability perspective,” McKain said.
“We have to look at the entire supply chain. So we have to determine what went wrong, who made the part or component that contributed to the incident, how did the part get made and was it modified or adapted after it left the plant?” he added.
By examining the supply chain of a part that will be used in some type of autonomous function, McKain said it gives The Hartford the opportunity to identify potential risks. He described a scenario where a manufacturer made solenoid switches for hydraulic arms. The switches and arms were put into a wheelchair lift on the side of a school bus. One day, there was an accident with the lift that involved a student.
“The manufacturer of the switch had no idea about the end use case of their product,” McKain said. “That’s why we have to look at the supply chain. If we know the end use case of a specific part, we can work our way backwards and let the company know what’s likely to happen from a risk perspective.”
Using a car accident as an example, insurers historically looked at who was liable for the incident. Maybe one driver ran a red light or made an illegal turn. With auto manufacturers using more technology in their cars today, it can complicate the insurer’s role with claims.
“It comes down to who the creator is and who’s coming to the table with these things,” Zarkowsky said of companies creating parts or components that are used for autonomous functions.
The Hartford’s Autonomous Technology Working Group
The Hartford’s Autonomous Technology Working Group is working to identify other potential issues with autonomy and hoping to find ways to solve for them. The Autonomous Technology Working Group was created after Zarkowsky noticed that the autonomous industry was becoming a larger part of the working economy.
“These companies are going from research and development testing to prototyping and developing products. And it’s not just cars, but other industries, too,” Zarkowsky said. “From my perspective, the industry is going from the small company stage to a maturation where they’re going to get bigger and get into true products.”
The goal, Zarkowsky added, is to make sure the insurer is prepared for autonomous systems as more businesses use them.
“One thing that gives me comfort is that we’re still using the basic principles of underwriting,” he said. “Instead of jumping in fully, we’re taking small, incremental steps. And what we’re doing is learning through those incremental changes to hopefully get us ready for when things get off the ground.”
One thing that the group realized was how insurance underwriting will happen on a “risk spectrum,” Zarkowsky said.
“You won’t say that because a business is using autonomous systems, there’s a hazard,” he explained. “You look at it as a spectrum. Maybe an autonomous device isn’t being used on public roads, but it’s in a warehouse. That doesn’t mean there’s no risk. There’s just a different risk than driving on a public road.”
How Does Insurance Coverage Change When It Comes to Autonomy?
Businesses using autonomous functions or planning to implement autonomy in their operations should review their insurance coverage. Despite the benefits it can bring, it also inherently increases risks in other areas, Zarkowsky said.
Some types of insurance to consider include:
- Cyber insurance
- Technology errors and omissions insurance
- Property insurance
- Products liability insurance
- Commercial auto insurance
- General liability insurance
What Will Autonomy Look Like in the Future?
It’s impossible to predict what the future will look like. But professionals at The Hartford believe autonomy will continue to rapidly evolve.
“If you look at where we were five to ten years ago, a lot has changed with respect to technology and autonomy,” Zarkowsky said.
When talking about technology’s evolution, McKain looked at the developments since the ‘80s.
“There used to be only one or two computers for an entire business. Now, we all have computers in our pockets that can do thousands of times that work,” he noted. “Technology is skyrocketing out there.”
When people think about autonomy, a common image is self-driving cars that know when to stop at a light or at a pedestrian crossing, while also being able to navigate safely around other cars on the road. McKain cautioned that for this to happen, it’ll take time.
“Technology has to develop all in concert,” he explained. “You’re talking about car manufacturers and tech companies creating products that can communicate with other businesses on the road and with municipalities. It’s going to take a while for all of these elements to align.”
Gluch emphasized the importance of maintaining principles despite a potential industry shift due to autonomy’s growth.
“Even though the technology may be different and it changes rapidly, a lot of the core underwriting issues and claims principles remain,” Gluch explained. “We still have to look at liability issues and adhere to disciplined underwriting standards.”