If you aren't already sold on the idea of autonomous vehicles imagine never hearing honking car horns. The computers will communicate with each other wirelessly. How pleasant.
Will autonomous vehicles use horns to signal human drivers who are about to cause an accident? Even if they do horn use will decline.
Autonomous vehicle research has turned up some lessons about dangerous human drivers. What struck me about it: if vehicles could just detect when drivers aren't paying attention the vehicles could apply the brakes and beep at the driver with an inside speaker.
Anyone want to hazard a prediction in the comments? In what year will vehicles under autonomous operation do 1% of the miles traveled per year? When will they hit 50% of the miles traveled?
Note how I phrased the questions, in terms of miles traveled, not in terms of vehicles with autonomous capabilities. The vehicles will not start out with autonomous capabilities in all driving conditions (e.g. highway operation will come first) and drivers will not activate autonomous capabilities at all possible times.
When making your prediction keep in mind that new cars account for just a small portion of total miles driven. The average car is kept for Cars are being kept for about 11 years (and here too) and reaching 150k+ miles. In recent years the average time owning a car has risen sharply. This slows the spread of a new technology on the road. Be very aware that newer cars are much safer.
On the other hand, imagine your job required you to drive 25k miles per year. Would you jump at the chance to get an autonomous vehicle and free yourself from clutching the steering wheel for hours every day? Early adopters might skew toward longer distance drivers.
Autonomous vehicles will cut vehicle accident fatality rates. Already many driver assist technologies in vehicles are contributing to a sharp decline in vehicle fatalities.
WASHINGTON – The U.S. Department of Transportation's National Highway Traffic Safety Administration (NHTSA) today released the 2013 Fatality Analysis Reporting System (FARS) data that shows a 3.1 percent decrease from the previous year and a nearly 25 percent decline in overall highway deaths since 2004. In 2013, 32,719 people died in traffic crashes. The estimated number of people injured in crashes also declined by 2.1 percent.
Longer lasting cars actually raise vehicle fatalities. A car built in the year 2000 is much more dangerous than a car built in 2015. What we need: upgrades for car computer equipment. This is already obvious in the passenger compartment of cars with ancient radios and other electronics. We also need upgrades for safety equipment so that one can add lane departure warning, collision avoidance automated braking, and the like.
How much does computer assist already reduce accident rates? Cars with front crash prevention tech have 7-15% lower insurance claims filed on them.
KPMG research shows that automation of some parts of the driving task have already led “to reductions in claim frequency.” According to the report “vehicles equipped with front crash prevention technology have a 7–15 percent lower claim frequency…[than] vehicles without it.” As portfolios of new products come to market “the downward drop in frequency will likely only continue.”
Revenue in the car insurance business is going to plummet along with declining accident rates.
Here is the full KPMG report. They are expecting about a halving of the accident rate from 2013 to 2027 or 2028. That is partly driven by the spread of technologies (e.g. front crash prevention systems) already going into some new vehicles today.
The incremental nature of a succession of driving assist technologies makes it hard to state clearly when autonomous vehicles will become available. Different people will take autonomous vehicle to mean different things. First, technologies short of full autonomy providfe successively more help to drivers who still have their hands on the wheel. Second, even when the autonomous systems start taking over fully they won't do it for all driving conditions. There is separated highway vs non-separated highway vs street in dense city vs street in suburb and so on. Also, there are a variety of weather and construction and accident conditions where many autonomous systems will ask the human driver to take over.
If we are going to argue about dates for various levels of autonomous vehicle penetration then for the foreseeable future we have to talk about how many miles driven autonomously rather than how many 100% autonomous vehicles are on the road. Autonomous technologies will take over some pretty high double digit percentage of total miles driven before the tech becomes mature enough that a blind person will be to travel alone in a vehicle anywhere a seeing person can drive.
|Share |||Randall Parker, 2015 June 20 11:59 AM|