Will autonomous vehicles drive into a safer future?

People are naturally afraid of change. When that change comes in the form of potentially relinquishing control over a basic function like driving, fear is accompanied by skepticism. For nearly every technological advancement in history — be it the electrification of homes, automobiles or space flight — reasons have been found why the change involves risk that should not be taken. In the end, however, all of these technological advances have produced benefits that have made us more comfortable, efficient and safe.

The accepted SAE International classification system for automated vehicles contains six levels from 0 (no automation) to 5 (fully autonomous). With the exception of testing environments, there are no vehicles on the road today that exceed Level 2. Level 3 vehicles may hit the road in late 2018.

Related:  Driverless cars may cut traffic jams, not insurance premiums

Defining levels of automation

To gain trust in the technology, it is important to distinguish between automated (levels 1-3) and autonomous or driverless vehicles (levels 4-5). The basic difference is the degree of human intervention.

An automated car has features like lane assist, automatic braking and speed control. However, the driver must be prepared to take over control in specific situations. A truly autonomous car would make decisions about lane change, braking and route and execute those decisions. At level 4, that would be limited to certain areas and situations; at level 5, there are no on-road exceptions, and no human intervention is required.

Driverless and automated vehicles hold the promise of a landscape where traffic congestion is minimized, stress is down, and accident frequency and severity are drastically reduced.

Eventually you will be able to ask Alexa to bring the car around and plan the most efficient route to work, while you relax and complete your Sudoku puzzle on the drive. That promise has not been fully realized yet. We have, however, already seen some very successful uses of autonomous cars across America — and some accidents, as well.

Naturally, this has led to some uneasiness. Adaptation to large and even midsize cities, outside of gated test areas, poses several challenges, but for better or worse, autonomous cars will become a part of our future reality. Public entities — including counties, cities, schools and anywhere autonomous vehicles will be integrated into the traffic system — should take steps now to pave the way for this future.

Related:  Liability in the age of autonomous vehicles

Challenges and risks

One reason some people find driverless cars scary is that they don’t fully understand the technology. How does the vehicle recognize the speed and direction of oncoming traffic? How does it distinguish between a pedestrian, a bicycle or a tree? One of the first concepts to understand is that driverless cars are being designed to be inherently risk-averse.

According to a rough estimate by McKinsey, driverless vehicles could actually lower traffic deaths by up to 90%. They are safer than having a person behind the wheel because, unlike humans, driverless cars act more predictably and don’t operate against their own interests.

They follow the speed limit — they understand what the speed limit is in a certain area, and they abide by it. They brake more quickly, and they have a faster rate of analysis than a human brain. They don’t get distracted by passengers texting or playing with the radio dial or kids screaming in the back seat. They can interact with other autonomous and even non-autonomous vehicles (with after-market vehicle-to-vehicle hardware and software installed) and accurately anticipate what they will do based on patterns and data from other vehicles.

Autonomous vehicles cannot predict what a person will do, however, because humans are unpredictable. Have you ever seen someone play chicken with a car at a crosswalk as the light changes from yellow to red, or gesture with their hand for a car to go instead of crossing? These sorts of variables pose risks – in Arizona, we recently saw a fatality involving an autonomous test vehicle and a pedestrian. The preliminary report from the National Transportation Safety Board provides data suggesting significant failures in software modifications, testing protocols, road design and pedestrian judgment.

Cyberattacks pose an additional risk. Recently, security experts at a technology conference got everyone’s attention by using a computer to make a Toyota Prius honk its horn and slam on the brakes. This is certainly terrifying in a world where cyberattacks are becoming more and more prevalent. For example, cybercriminals could hijack a traffic control system and send incorrect information to vehicles, resulting in potentially catastrophic injuries. Vehicle manufacturers and public entities will need to build trust by informing consumers and constituents of the security precautions being taken while being fully transparent about the residual risk that will still be present.

Related:  Expect a shift in the auto insurance litigation landscape

Insurance issues and solutions

In terms of insurance, there are several factors at play. As an emerging technology, autonomous vehicles necessitate development of coverage solutions and evolution in legal theories of negligence. Reduced traffic and the subsequent decrease in property damage, bodily injury and associated claim expenses will lower the overall cost of claims.

However, this could be a double-edged sword, as the potential arises for increased product liability costs. Was it a software or hardware malfunction? Did the consumer alter the software or fail to install required updates? These sorts of questions will be asked as insurance policies and case law catch up with the technology.

As with any risk assessment, some of the first steps will be to determine the extent of the risk as well as the tolerance for that risk. Will companies assume the risk, retain it, or transfer via contract to an insurance company? Insurance carriers need to develop their understanding of the technology and how the extent of deployment and different levels of fleet automation could require changes in policy language, the reevaluation of loss trends, and affect an entity’s rates.

How to deal with autonomous vehicles

City infrastructure needs to connect with the autonomous system to allow for better and faster movement of vehicles. For example, smart streetlights, traffic signals, sidewalks and parking garages should be set up to maximize the throughput of driverless vehicles. Local governments should actively encourage investment in vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) technology, which can make driverless and highly automated cars safer. Examples include federal Smart City grants, matching loans from state infrastructure banks or public/private partnerships to accelerate investment.

In December 2016, Audi launched the first V2I technology in the United States, in Las Vegas. The technology sends real-time signal information via an onboard cellular data connection from smart traffic management systems that monitor traffic signals. In addition, big data collected from smart garages and streets can accurately predict traffic patterns and facilitate faster commute times. Many other cities and university campuses such as Atlanta, Boston, Kansas City and the University of Michigan’s Mcity have followed suit.

The goal is to gather data on how driverless cars interact both with other vehicles and a city’s infrastructure and learn what is needed to deploy the technology fully.

Related to V2I technology is the need for better-made and easier-to-understand traffic signs, lines and infrastructure in general. Many cities rely on old, crumbling infrastructure and could use new connected signs, reactive paint and roadside-assist communication. Self-driving vehicles will rely on cameras, light sensors, GPS and other advanced technology to identify people, bicycles, signs and traffic control advisories on roads.

Vital to many of the V2V and V2I interactions are the speed, security and available spectrum of 5G networks. Reliable wireless communication is crucial to the functionality of autonomous vehicle technology and its integration with infrastructure. Municipalities should proactively reach out to regional wireless broadband providers to plan for related infrastructure growth so that future needs — such as public safety and transportation — can be fulfilled.

There is no hiding from cyber liability. In the event that self-driving cars fall victim to cyberattacks, municipalities must adjust their cyber, automobile, general liability, public officials and even law enforcement liability insurance coverage, accordingly. As with any new technology, the coverage implications will extend far beyond the obvious auto liability concerns. Municipalities must engage their broker, agent and carrier partners early in the process to address current exposures in testing environments, and start discussing the road map for future deployment of the technologies in real-world situations.

Safety in general must also be addressed. “Cities have an opportunity to come together and lobby their state governments to advance their concerns around the safe operation of autonomous vehicles in their communities,” according to the National League of Cities in its autonomous vehicles policy preparation guide. This means managing the risks proactively and in the same framework as the entity’s overall enterprise risk management.

Considering the aforementioned risks, now is the time for governmental entities to create policies specifically oriented toward autonomous vehicles and related technology. Government planners should determine how existing laws and municipal codes will influence the development of autonomous vehicle technology. That means working with urban planners, public works managers, IT specialists, law enforcement leaders, regional agencies and others to create a plan that addresses the deployment of this technology and its integration with existing transportation infrastructure.

Boston has already begun working on a long-term plan for autonomous car technology, which involves working with state transportation experts and partnering with the World Economic Forum to test autonomous cars.

Driverless cars are coming, but with the right planning and determination, major risks can be identified and addressed.

Thom Rickert (thom.rickert@tridentpublicrisk.com) is vice president, head of marketing and emerging risks specialist at Trident Public Risk Solutions, member of Argo Group. Rickert has extensive underwriting and marketing experience in all property and casualty lines of business, spanning multiple segments and industries, with a special focus on public entities and education.