How close are we to self-driving cars?
If you live in the Bay Area, odds are good you’ve seen a self-driving car on the streets—Google’s autonomous vehicles are already a fixture in Mountain View and its vicinities. Ford, Nissan and Tesla have all announced plans to introduce self-driving cars in the next few years and by 2030, up to 15% of new cars could be fully autonomous.
Transitioning from human-driven to self-driven vehicles
Self-driving cars face a number of hurdles in the short-term as we transition from a human-controlled fleet of vehicles to one that (at least theoretically) could become completely self-driven in the future. We'll examine each of these below:
Unfamiliarity with technology
The push for self-driving cars is largely motivated by a desire to increase safety. The evidence so far suggests that automated vehicles will decrease the overall number of accidents and mitigate the serverity of those that do occur. However, a recent survey by AAA found that nearly three-quarters of US drivers admitted that they would be afraid to ride in a self-driving vehicle. Interestingly, the same study found that drivers whose vehicles had semi-autonomous features such as automatic braking and lane departure warning were 75% more likely to trust the technology than those that do not. This suggests that as drivers become more familiar with how these features work, the concept of a wholly self-driving vehicle will be easier to accept.
Mix of human and autonomous drivers
As the technology becomes more pervasive, most experts agree that the majority of crashes involving self-driving vehicles will be caused by human error on the part of the other driver. Autonomous vehicles act in a predictable, law-abiding manner and could potentially communicate electronically with other self-driving vehicles to avoid crashes. Adding human drivers to the mix creates a layer of complexity that is harder to anticipate.
A study by the University of Michigan’s Transportation Research Institute study found driverless vehicles were most often hit from behind by inattentive drivers “unaccustomed to machine motorists that always follow the rules and proceed with caution.”
Partially engaged drivers
As cars become increasingly automated and distractions from cell phones, etc. become more wide-spread, drivers may have a harder time paying attention to the road when necessary.
Current autonomous features act as back up for drivers whose attention wanders or whose visibility is limited. If the human becomes the back up to a mostly self-driven vehicle, the odds of that human being actively alert at the critical moment could fall dramatically.
How autonomous are today’s vehicles?
Many of the features that power self-driving cars are already in the cars being sold today. The same AAA study cited above found that 61% of US drivers were interested in having semi-autonomous driving technology in their next vehicle including:
Forward collision warning and auto-braking
Forward collision warning systems use cameras, laser beams and/or radar to scan the road ahead and alert the driver to any objects in the road ahead. If the system detects an object that the driver does not appear to be reacting to it takes action.
Major automakers have agreed to make automatic emergency braking systems mandatory in most cars and light trucks by 2022. It’s estimated that this technology could avoid or mitigate more than half of the 1.7 million rear-end crashes that occur each year in the United States alone. (Should your next car have a crash prevention system?)
Adaptive cruise control
Adaptive cruise control uses a small radar (or laser) unit under the front grill or bumper that measures the distance to the vehicle in front of you. The system uses this information to calculate distance and speed of the vehicle ahead and react to any changes to maintain a safe driving gap. In the event the vehicle ahead brakes suddenly, the system will either alert the driver or, in some cases, apply the brakes to prevent an accident. (Should your new car have adaptive cruise control?)
Lane departure warning and prevention
Lane departure warning systems use cameras to detect the lane markings on the road. If the driver moves outside of the marked lanes without using the turn signal, an alert appears. Typically this is a visual alert combined with an audible tone or vibration. Lane departure prevention takes this one step further by gently steering the vehicle back into its lane. (Should your new car have a lane departure warning system?)
Blind spot monitoring
Blind spot monitoring systems use cameras or radar to watch the area around the vehicle and look for other vehicles that are nearby. In most systems, a light will appear in the side-view mirror to alert the driver that a vehicle is present. If the driver activates the turn signal while another vehicle is in the blind spot, an audible alert or steering wheel vibration is activated. In some systems the vehicle will even steer itself away from the approaching vehicle to prevent the accident. (Should your new car have blind spot monitoring?)
A $20,000 self-driving vehicle?
Self-driving features are no longer limited to expensive, well-equipped luxury vehicles. Honda recently released a Civic LX for as low as $20,400 that can drive itself on a highway today under most conditions. As long as the lane markings remain visible and another vehicle is in front of the car, the Honda can not only maintain speed, but swerve, slow down or hit the brakes as necessary.
It seems clear that, in the short term, we'll see an increase in the number of autonomous features being offered in vehicles. Whether or not this eventually leads to a world of self-driven vehicles has yet to be determined, but it's looking more likely.