How long before your car takes you for a ride?
That’s the million-dollar question, and despite all the predictions you’ve heard about autonomous cars, the actual answer is, “Nobody knows for sure.” Much of the technology already exists – you can currently buy vehicles that obediently stay in their lanes and follow traffic with your hands off the wheel, and some can even change lanes – but the bigger picture is considerably more difficult.
“There’s a common theme in the industry for automated driving that 95 per cent of it is pretty easy, and the remaining 5 per cent is really, really, underlined-really tough,” says Brett Smith. He’s the assistant director of MET (Manufacturing, Engineering & Technology) at the Center for Automotive Research in Ann Abor, Michigan. “I saw somebody saying (automated driving) will be in five or 75 years, and they don’t know which.”
From reactive to proactive
To hit the streets on their own, vehicles have to be proactive, and at the moment, the auto-driving cars available are only reactive. They only stop if something ahead stops, and stay in their lane only if road markings are visible.
A truly autonomous car will have to be able to do everything you do: know what colour the traffic light is, navigate a road with snow-covered lines, or realize a pedestrian is about to enter the crosswalk.
And that’s just the stuff on the street. A vehicle missing its steering wheel will have to find an empty spot in the parking garage, make its way to a refuelling station, and even get itself onto a hoist when it’s time to change a tire.
Mucho communication… and mucho ducats
All that will require vehicles to be far more communicative than they are right now. Part of it will be vehicle-to-vehicle, or V2V, when cars “talk” to each other to make their way around. They may send a signal when they’re approaching an intersection, for example, or tell cars behind them if there’s an incident ahead, which will be relayed down the line.
Another piece of the puzzle is vehicle-to-infrastructure, or V2I, and that’s the tougher one. The technology itself is relatively easy: traffic beacons to indicate road conditions, red lights that say stop, or parking lots that guide cars to the spaces.
But that’s a lot of investment, and if your municipality doesn’t even have the cash to fix all the potholes, who’s going to pay for the infrastructure?
“The real challenge for automated, connected vehicles is if they can move beyond a geo-fenced region,” Smith says. “A 50-block part of Toronto is doable but challenging, but then opening it outside of that is probably best described as that next 5 per cent. You have to build out to areas that you don’t think you have to build out. Take it into a rural area, and that becomes a bigger challenge.”
Big Brother… 2084
Another issue is accountability. “If you can get the technology to 99 per cent, you are better than a human driver, but you still have liability issues,” Smith says.
A major challenge is figuring out just who’s responsible when something goes wrong – and of course it will, whether it’s a glitch in the car’s software, or someone hacking into the highway system.
There are privacy issues: the infrastructure providers will know exactly where you are, every minute of the day.
Even seemingly minor questions need to be answered. Will you still need personal insurance? Will you need a driver’s license? Will children or an intoxicated person be permitted to ride alone?
Certainly you’ve read about the ethics – given only two choices in a situation, does the car kill the busload of children, or its driver?
And in the interim period, when self-driving cars share the road with conventional ones – and that could cover 15 years or more – how do they cope with vehicles that aren’t communicating with them, and doing things they don’t expect them to do?
Culture Crash – sorry: Clash
Making all of this tougher is what Smith calls a “culture clash” between the companies working on autonomous vehicles. “The auto industry has traditionally been extremely cautious and conservative, and for good reason and frustratingly so,” he says.
“They tend to realize that if you do something wrong, it could kill people and cost you an enormous amount of money. Now that conservative culture is being stood on its head by Silicon Valley, which says, ‘We’ll get it on the road, we’ll beta test it, if it doesn’t work we’ll say oops, and if someone sues us, we’ll close and move down the street, and if the investors get burned that’s how it goes.’ One isn’t bad or good, it’s just different ways of thinking.
“The high-tech, high-risk, high-investment thought process is changing the way that traditional car companies have to react to this process. Tesla is great for this industry because of how it can change the industry, but it’s (also proof) of how tough it is to build a car. It’s a fascinating change, but I don’t think it’ll happen as quickly as some of the technology company forecasts will tell you it is.”