Editor’s note: To hear the full interview click on the podcast above or visit www.insuarncejournal.tv.
Hilary Rowen, a partner in the law firm of Sedgwick LLP in Sacramento, Calif., for the last three years, has been analyzing and following the insurance issues relating to the emergence of autonomous vehicles, particularly self-driving cars.
She has a lot to say about the insurance and legal implications of letting go of the steering wheel and handing over control to a computer.
Rowen, who also has some opinions on new regulations from the California Department of Motor Vehicles, spoke with ÈȵãºÚÁÏ Journal about these matters.
ÈȵãºÚÁÏ Journal: What are the key provisions of the new California DMV regulations on self-driving cars?
Rowen: The new regulations just issued last month cover operations by consumers – ordinary drivers who purchase cars, not just the testing of vehicles. The key features of interest to the motoring public, as well as to the insurance industry, are the DMV regulation only permits leasing by auto manufacturers and only for a period of three years.
They require third-party testing of cars, which is new. The National Transportation Highway Safety Administration, NHTSA, basically allows self-certification by manufacturers. In addition, the new DMV regs require a separate driver certificate, not a separate license, but a separate certificate before someone can be behind the wheel of a self-driving car.
The regulations require that the car be capable of disengaging when it hits conditions that it cannot handle. In other words, if the car hits heavy rain, snow, other road conditions where the self-driving features don’t work, it has to be able to automatically disengage and hand control back to the human operator.
ÈȵãºÚÁÏ Journal: How do these regulations compare with what’s happening in other states and nationally?
Rowen: These regulations are far more detailed and specific than what’s happening in other states. Most states have not promulgated regulations or statutes dealing with self-driving cars at all. A few states have determined that their motor vehicle codes, in fact, already permit the operation of self-driving cars and those legislatures have not to-date taken any action to add provisions relating to self-driving cars.
California has a much more detailed regimen, very much focused on basically safety considerations. It also basically, currently does not even permit the testing of completely control freed self-driving cars. In other words, you just get in the car, you push in the destination, you hit the “Go” button, unless the test vehicle is retrofitted to add operator control.
There are some concerns that testing may be moved to other states and whether this will have an impact on whether California remains a design center for self-driving cars.
ÈȵãºÚÁÏ Journal: Are self-driving cars likely to change car ownership patterns?
Rowen: I think in the long term, yes, and indeed the DMV (regulations), by mandating that at least for … three years following the actual approval for a specific make and model to be operated on the California roads that the vehicles be leased by the manufacturer only.
Even before that, a number of commentators had indeed suggested that there may be a significant move to having self-driving cars be either fleet vehicles or vehicles that are insured by the manufacturer, thereby effectively shifting a lot of the premium into either self-insurance programs or in to commercial auto.
Auto owners or auto users, I’ll call it auto leasers, would still have to maintain an auto private passenger auto policy, but it may turn out to be a much lower premium policy with the resulting reduction in commission income for agents.
ÈȵãºÚÁÏ Journal: We’ve heard a lot about is the shift in the liability from an auto product to a liability coverage. Do you see that happening?
Rowen: I think it will happen. I think, however, you’re going to see an intervening period in which there is a lots of dispute on accident causation patterns, particularly when you’re talking about vehicles where the human driver can retake control.
There’ll be numerous disputes about whether a particular accident was caused by the failure of the technology and therefore be a product liability claim, or by the driver imprudently taking back control and therefore causing an accident, thereby being a private passenger auto claim.
ÈȵãºÚÁÏ Journal: How will insurers rate these vehicles?
Rowen: That is going to be a really interesting question. It’s going to be a particularly interesting question for two reasons. One is, we are now seeing cars where effectively the driving risk characteristics can change due to software updates. The original make and model may no longer be the basis for developing statistically significant samples data for determining the risk of a given make and model.
I know from talking to auto insurers, they are quite concerned about how they can stay on top of the nature of the autonomous technology included with the car when it sold and how that technology or software capabilities are going to change during the life cycle of the car.
Second reason is that, at least in California where we have Proposition 103 that mandates that the greatest weight be given to the driver’s driving record, it’s unclear how you rate a vehicle where the car and not a human being is doing the driving.
ÈȵãºÚÁÏ Journal: Will this all reduce auto insurance premiums, or will it increase them?
Rowen: The general view, and certainly my view is that, particularly as the fleet increasingly moves towards more autonomous features is that BI rates will go down, bodily injury rates will go down – both frequency of accidents and the severity of the accidents. In all likelihood the frequency of property damage claims will also go down.
It is likely that the severity, the cost per claim on property damage, will go up because cars will now have very expensive sensors hung in, or arrayed in, very vulnerable locations. I think net-net, you’re going to see a downward trend in on auto insurance premiums, but you’re likely to see this will take a while, partly because the fleet only turns over slowly. That is you can introduce an autonomous vehicle in 2020, and you’ll still have a significant number of conventional cars on the road in 2030.
Related:
- Crash Rates Lower for Self–Driving Cars: Virginia Tech Study
- Federal Transportation Officials Want Self–Driving Cars Sooner
- California DMV’s Pedestrian Pace on Self-Driving Car Rules Makes Google Antsy
- Makers of Self-Driving Cars Could Shoulder Burden of ÈȵãºÚÁÏ Liability
- Automakers Outpacing Tech Firms in Driverless Car Patents
- Ethics of Self–Driving Cars: Should They Kill Driver to Save Others?
Topics California Auto Legislation Personal Auto Manufacturing
Was this article valuable?
Here are more articles you may enjoy.