Think about how many human skills have been lost in the last century. It seems we are becoming less skilful and relying on experts to do things for us. A local business charges hundreds of dollars to service a bicycle. NRMA comes to change a tyre. Will we resist driverless cars because we don’t want to lose yet another skill? And what would we do instead – watch more movies?
Underestimating the skills of humans
It’s easy to underestimate the scale of human accomplishment. Auto and tech companies seem to have underestimated the skills involved in driving car. They also overestimated their skills in building an autonomous car that is as clever as a human.
As Markus Schäfer, Mercedes said, “Now we understand what a human being is capable to do and to imitate that is a huge undertaking”. This statement followed 6 years of work across two continents and billions of euros.
Driving the average manual car used to demand 1,500 to 2,500 different skills. Even with assisted driving tech, like automated parking, we still need hundreds of skills to manage the surprises in each trip.
Experts call these edge and corner cases – things that don’t happen that often. Developers of autonomous cars must build in the certainties we know about while driving plus the uncertainty of things we don’t know we don’t know.
Overestimating what cars can do
Many drivers with assisted driving features overestimate what their cars can do. According to Thatcham, this is just the first of three phases of driving: assisted, automated (or semi-autonomous) and autonomous.
For example, assisted driving is where the driver has full responsibility but shares control with the vehicle. Autonomous driving is where the vehicle has full responsibility for control and the driver is just a passenger. The transition from assisted to autonomous is huge and loaded with unknowns. This is where the most danger lies because it needs to be clear exactly who, or what, is doing the driving.
We may soon have a potent mix where some vehicles have no safety technology, others are semi-autonomous and a few fully autonomous vehicles are being tested. Any car in a mixed environment must be able to predict what other drivers, pedestrians and cyclists might do.
How can drivers react to another vehicle if they don’t know who is driving? Worse, drivers may stop feeling in control of their own vehicles, even when they are.
Underestimating their safety
How can we be sure autonomous vehicles will be safer? There is currently no standard definition of safety as it applies to AVs. Any argument that AVs will lower deaths and injuries from car crashes can be crushed by a single, dramatic AV crash.
Professor Graham Currie, Monash University claims: “We are at the top of the hype curve at present, and much of what is being said about driverless car performance is overly ambitious speculation.”
Although AVs must be safer than conventional vehicles, the true impact of AVs is still unknown. To prove AVs outperform human skill demands an accepted, watertight method of keeping score. There isn’t time to drive tens of millions of kilometres to create statistically significant volumes of crash data.
A recent patent application in the US measures the safety of AVS by comparing behaviours, such as hard braking, sharp turns and sudden acceleration, with standard cars on the same stretch of road. Rand Corp suggests AVs need leading indicators for measuring safety, not lagging ones, such as crashes. It proposes the idea of roadmanship, which measures whether a vehicle “plays well with others” in traffic.
Meanwhile, legislation is still lagging when it needs to be leading. National Transport Commission (NTC) is still seeking submissions to an AV inquiry and will not even present findings to transport ministers until 2020.
Overestimating demand for autonomous cars
It is difficult to measure demand for AVs when they do not currently exist. But researchers are constantly asking about intentions.
An international study of 1,500 drivers in Australia, France and Sweden found French drivers were keener on AVs than those in Australia and Sweden. In France and Sweden, people’s emotions and beliefs about these vehicles were the biggest predictor of acceptance, compared to Australia, where it was expectations of performance. Perhaps Australians wanted to be sure the car would drive with as much skill as they do!
A small Queensland study asked drivers about the advantages and disadvantages of AVs. Typical answers were given – freedom to multi-task, reduction in road trauma – and the disadvantages of technology malfunctions, legal liability, lack of trust and control. They also didn’t want to lose the enjoyment of driving a car themselves.
Meanwhile, how much will an AV cost? Many people say it will not be possible for most individuals to buy them. Only big corporates and fleet owners will be able to pay what these vehicles are actually worth.
Underestimating the role of insurance
Without a driver, there is no driver to insure. Premiums depend on your likelihood of being in an accident, as well as actual crash rates in your area. With AVs, there is no history of crash rates, and no driver history to be had. So why would we need auto insurance?
Data show 90% of crashes involve some kind of human error. But if 90% of AV crashes involve some kind of machine error, what is the difference? There is still a crash, and someone could be hurt or killed. Moreover, AVs would be extremely expensive to repair, with all the sensors and software involved.
Policies that protect products, rather than people will become more widespread. If mobility as a service becomes standard, we might insure our safety as a passenger rather than a driver.
Meanwhile, most drivers continue to drive with skill and without incident. One academic suggests the chances of a human driver having a crash, compared to not having a crash, could be one in a billion. You would need to measure how many times in the year you did not have a crash, compared to how many cars you could have crashed with, and how long you spent driving on the roads.
Are we really willing to lose our driving skills? Perhaps we would rather drive ourselves.