Though once firmly rooted in science fiction, driverless cars are quickly becoming a reality. Around this time last year, Uber was testing self-driving Uber cars in Pittsburgh and reports of Google, Tesla, and many other top tech developers working on similar projects could be found across the internet. It’s a year later, and we are quickly approaching the earliest estimations for when these vehicles will hit the market, and yet as The Economist points out, there are a number of roadblocks still in the way before driverless cars can really hit the road.
- Litigation: Even after a new technology finds market success, there is still a long process of adoption. Different people will purchase a new technology at different times. For driverless cars, that means our roads will be a mix of A.I controlled vehicles and human controlled vehicles. When there is an accident between these two types of vehicles, who is at fault? The owner of the automatic car? The human driver? The manufacturer? New policies and regulations will need to put in place to determine the legalities surrounding driverless cars.
- Test Drives: Knowing whether or not these cars are safe is a much more difficult process than one might assume, and it all comes down to math. The Economist states that there are about 1.12 deaths and 76 injuries per 100 million miles driven across the country. In order to be certain that the driverless car software is truly safe, these test fleets need to drive hundreds of millions or billions of miles to prove they are safe. Furthermore, there are disputes over how safe these cars actually need to be, so the standard is likely to shift as these technologies come closer to market.
- Different Levels of Automation: When people think of driverless cars, they assume a certain level of automation. They often assume we are discussing completely autonomous vehicles that transport people to specific destinations with no control from the driver. That’s actually known as level 5 automation and it represents the highest tier of the scale, and there are arguments to be made that certain lower levels of automation are actually less safe than no automation at all. Level 3 automation only requires the human passenger to take control of the vehicle in emergencies, but that can lead to drivers failing to respond appropriately in emergencies due to lack of experience. There will likely be pressure for these developers to advance beyond this level before anything reaches the public at all.
- Infrastructure: Local governments are going to need to prepare the roads for driverless cars in ways we may not have anticipated. Vehicle to vehicle and vehicle to infrastructure wireless networks are going to need to be built on the roads to ensure these driverless cars can navigate traffic safely.
If Not Now, When?
Some estimates place driverless cars on the road in as little as two years, but given the challenges above, two years seems like a longshot. Other estimates say 2021, and that may be more reasonable. The real takeaway is that the driverless cars we see first, will not be the level 5 autonomous cars we assume. They are likely to still involve some human control or at least offer the option. None of this changes the fact that this technology is coming and coming soon. It may just be a tad longer than we anticipated.
For Further Reading: