About 250 million passenger vehicles crowd American roads, according to the U.S. Bureau of Transportation statistics. They take different paths, yet new models move ever closer to a common goal: autonomy. This infographic by AutoInsurance.US shows that Google has been out front in the autonomous vehicle effort, modifying a fleet of cars – mostly Toyota Priuses – with self-driving technology and testing them. It plans to take its technology to market in 2018.
But are drivers – and states – ready for it? Maybe not. Only about 17.8% of respondents to the AutoInsurance.us poll said they would buy a self-driving car. And state legislatures are only beginning to consider how to regulate the technology.
Stanford University's Center for Internet and Society monitors the progress states have made formulating, considering, and passing autonomous-driving legislation. So far, California, Florida, Nevada, and the District of Colombia have adopted measures. But they've mostly stuck to defining autonomous vehicles and directing their DMVs to create rules for them before 2015.
Florida, California, and the District of Colombia contend that a licensed driver must be behind the wheel of any autonomous vehicle, prepared to take control at any moment. So has Florida, though it has exempted drivers of autonomous cars from a ban on texting and driving. Nevada has a similar regulation but adds that a driver of an autonomous car must never "drive" drunk.
As of August 2013, eight other states were considering new autonomous driving measures. Again, most bills being considered direct local DMVs to create rules for the road by 2015, but there are some individual state additions. Hawaii's bill, for example, would exempt a non-reckless operator of an autonomous vehicle from liability in a wreck.
If bills similar to Hawaii's are passed there and in other states, it could certainly shake up the insurance world, along with the concept of driver responsibility in general. To understand the risks involved with liability, however, you must first comprehend the different levels of autonomous vehicles.
Hands-free in five steps
Autonomous-vehicle technology experts use the German Federal Highway Research Institute's system of categorization for the different levels of vehicle autonomy. The classifications are as follows:
Level 1: All technology must be operated by a licensed and alert driver. There is no automated assistance in this type of vehicle.
Level 2: Vehicles contain some driver-assistance technology. The driver permanently controls either longitudinal control (cruise control/emergency breaking) or lateral control (lane guidance), and the other task may be automated to a certain extent. These types of vehicles already frequent highways. The most widely used level 2 technology is cruise control.
Level 3: These are partially automated. The system takes both longitudinal and lateral control. The driver permanently monitors the system and must be prepared to take control at any time. Level 3 vehicles, though not as prevalent as level 2 vehicles, already are on the road.
Level 4: These vehicles are highly automated. Google's test fleet falls under this group. The vehicle's system takes longitudinal and lateral control. The driver no longer is required to permanently monitor the system but must be prepared to take control within a specified time limit in the event of a takeover request.
As technology developers test these cars, they must consider ways to keep human drivers engaged so they will respond when the system calls them to action. With this level of automation, the driver will not be able to sleep, which is the main draw for 31.9% of consumers who said they would buy a self-driving car, according to the AutoInsurance.us study.
Level 5: These cars are fully automated. The vehicle controls every aspect of driving. If a system's takeover request isn't answered, it can revert back to a minimal risk condition. A driver's attention would no longer be required at any point. Only this level of technology would allow people to sleep or place children alone in a moving car.
The leap between levels 4 and 5 is complex. It requires a vehicle to possess enough logic to recognize its own technological failures and correct them. Level 5 systems also must be able to answer complicated, morally ambiguous questions in emergency situations, such as weighing potential injuries to passengers vs. injuries to pedestrians in the event of an unavoidable accident.
The new blame game
When level 4 cars make their way onto U.S. roads, liability issues could get tricky. Who's at fault when a level 4 car crashes, the company that developed and built the autonomous technology or the negligent driver of the vehicle? What happens if an autonomous car crashes into a non-autonomous vehicle? How will the word of a driver stack up against the word of a technology giant such as Google?
Once the technology is perfected, will there be any need for auto liability coverage, currently required in every state? Or will there always be just enough doubt to make it necessary?
Running over our own privacy
Controversy frequently breaks out in the U.S. about privacy. With advent of self-driving cars, those concerns could grow. Many motorists provide locations and destinations to the GPS systems in their cars and phones without giving it a second thought. But will they be as willing to provide such information to autonomous vehicles, which must coordinate trips with millions of other cars on the road. Will the information be centralized? Who will be able to access it? How secure will it be?
Questions about liability and privacy are just two examples of the ways in which consumers should consider the whole picture of autonomous driving. Other concerns could involve speed limits, ticketing, and the ecological effects of vehicle automation. It's easy to get caught up in the glamour of autonomous cars, but will there be unintended consequences that undercut the convenience?