- About Us
- Exclusive Programs
- Specialty Programs
April 8th, 2017
From acquisitions to self-driving test drives to insurance responsibility to regulations:
With the addition of Mobileye, a leader in collision prevention sensors, Intel gains technology that “uses artificial vision sensors placed in the cars, which act like a driver’s third eye to alert drivers if a pedestrian is in the road, or if they’re approaching another car too quickly.”
Intel CEO Brian Krzanich explains, “Mobileye will provide ‘eyes’ that helps self-driving cars ‘see’ the world, while Intel will provide the ‘brains’ that plot a car’s path. Uniting these two technologies into one platform will make it easier for lawmakers to regulate the technology.” Krzanich expects self-driving cars on the road by 2023.
But that day could arrive sooner.
In March, Uber launched their self-driving car tests in several cities but it was not without hiccups. A test car in San Francisco, caught on traffic video running a red light, sparked a media debate on the validity and safety of self-driving cars. In Phoenix, an Uber in self-driving mode was flipped onto its side after another car (powered by a human driver) “failed to yield” to the Uber at a left turn. Nobody was injured and the Uber car was deemed not at fault.
In both instances, the argument can be made that these self-driving Ubers aren’t fully autonomous, as they were each operating with a human “safety” driver who had manual override capability. Uber did pull their self-driving cars from three cities after the Phoenix crash, but the program resumed days later.
While fully autonomous technology has not been publicly rolled out on a large scale, Tesla’s Model S boasts an autopilot system that’s considered a near-full autonomous driving experience (though the company does remind drivers to keep their hands on the wheel).
In late March the company pushed out their new software update said to improve the Autosteer, Auto Lane Change, Lane Departure, and Summons functions, bringing the driving experience closer to fully autonomous.
Several automakers currently sell models with auto-functions available; self-parking, lane-keep assist, adaptive cruise control, auto-braking, but the percentage of cars on the road utilizing multiple auto features is small and with none fully autonomous, that leaves regulations, safety, and responsibility, highly debated topics.
Will the car owner, the auto-maker or the software creator be held responsible in the event of an accident or driving infraction?
Warren Buffet, head of a conglomerate Berkshire Hathaway which owns commercial auto insurance subsidiary Geico, claims self-driving cars “could become a major threat to insurance companies when the technology hits the market. If autonomous vehicles prove to be safer than regular cars, insurance costs will plummet…if they’re safer, there’s less in the way of insurance costs, [and] that brings down premium buy significantly.”
How this technology will affect the insurance market is a consideration, prompting insurance representatives, when meeting with the California State Senate Committee on Insurance, to request more access to autonomous vehicle data.
A lawyer representing a large insurance firm claims, “Data access is essential to develop proper pricing and underwriting of vehicles, critical for liability determinations and from the general public’s perspective, important in determining the safety and reliability of the technology.” He further explains, “underwriters are not seeking information on every aspect of [a] vehicle. But they do want data that shows driverless-car technology actually does what it advertises it’s doing.”
Which is a similar stance to that of the California DMV, from a regulations and safety point of view. Chief counsel Brian Soublet says, “if you program your vehicle to obey the law, it’s not going to run over pedestrians, or crash into other vehicles. I don’t want to say we’re comfortable. We believe we’re requiring certification from the manufacturers that they’re ready and that the vehicles themselves are able to operate without causing some harm.”
New regulations proposed by California DMV in March cover aspects of public safety while working to promote innovation, testing, and deployment of autonomous driving, to accommodate the large number of car manufacturers testing the technology in the state.
California “has more manufacturers testing autonomous vehicles than any other state”, according to California Transportation Agency Secretary Brian P. Kelly.
In fact, just this week, the DMV granted approval to Apple to begin testing, bringing the total companies with permits up to 30.
California’s concession to test truly driverless cars, with no human driver assisting, allows for vehicles with no steering wheel or pedals on the road, but currently most testing is being done with human drivers. A public hearing is planned for April, to discuss the proposed changes and new regulations.
It’s clear laying the groundwork for regulations and data mining to gauge costs and safety are the necessary next steps, but it’s early in the development to really know how autonomous cars will perform, and if they will perform better than humans.
Which means pinpointing where insurance responsibilities lie, with car manufacturer, software creator, or consumer, is still a gray area.
Currently, insurance rates are calculated by multiple factors; driving record of the listed drivers, value of the car, type of car (newer cars cost more than older cars), miles per year driven, usage (pleasure/business, long and or short commute), limits selected, deductibles, where it is garaged. Additionally, more expensive body work to repair vehicles from complex electronics and cameras can drive up rates.
But with self-driven cars on the road, equipped with expensive technology, the current calculations become null.
Until more data is available, “fault and liability will be determined case by case, as happens after accidents now”, states Bryant Walker Smith, law professor at the University of South Carolina.
James Lynch, vice president of data and information services and chief actuary of the New York-based Insurance Information Institute has a similar outlook, “what it depends on is who the courts or the adjudication process finds to be at fault. It could be with the manufacturer if there are defective parts. It could be with the state if they didn’t update their signs, or it could be with the driver if they didn’t download an update they needed to. But whoever the liability sits with, they’re probably going to have insurance.”
Some autonomous auto enthusiasts are optimistic that self-driving cars will eventually outperform humans. Hod Lipson, professor of mechanical engineering at Columbia University claims, “driverless cars keep getting better the more they drive, whereas humans have a roughly constant safety record over the years. The idea that somehow a human driver makes the drive more secure is false comfort, and potentially dangerously misleading.”
The truest safety data will come when autonomous functionality is standard in all vehicles. Until multiple self-driven cars are on the road, at the same time, it’s difficult to calculate the differences between a car’s reaction and a human one.
But with companies like Intel propelling driverless technology forward, car manufacturers adding features and ramping up testing, and lawmakers imposing new regulations, it will happen. It’s time for consumers and the Insurance industry to ready for change.
The reality of fully autonomous driving cars is no longer theoretical. The future is here.