I recently had to change a newspaper delivery schedule by phone, which meant talking to a computer (in vain) for 10 minutes before finally getting through to a human operator who understood my request. This got me thinking about the frustrating nature of computer interfaces, which is a major obstacle to public acceptance of self-driving cars.
To find out how people willl likely respond to the new technology, I conducted a small informal survey among acquaintances. I asked people of all ages whether the totally self-driving car (no familiar steering wheel) was something that excited them. The common response from the under-40 age group was “yes” and the reason given was that such cars would eliminate the danger of reckless drivers who commonly cause accidents. They did not, however, express concern about the computer interface.
The older group’s response was nuanced – they did not want to appear scared of new technology but were not sure what it would mean for them. I explained to them that the truly self-driving cars promised in the distant future would be run by computers always connected to a wireless network and that such cars would have to be programmed by passengers when starting a trip or when changing itineraries.
Few exhibited any enthusiasm for such a car for their personal use, although they did not doubt that it would be used for certain commercial services.
Digging deeper into the responses, it became clear that the believers had faith in the ability of computers to exercise better judgement than humans in difficult traffic situations. They have also come to believe that computer interfaces are painless.
When I pointed out that the driving computers would have to have access to enormous visual data accessed through an always-on wireless connection that would guide the actions of the car, most people did not seem to consider that a potential barrier: Their answer was that the technology will eventually be there to do just that.
However, when asked whether they wanted to be in the position of frequently reprogramming the computer while driving as they change their itinerary, sometimes with a destination without a a computer-ready addres, they became hesitant.
Historically, successful services or products provided via computer applications for mass use have been user-friendly enough to make consumers forget that a computer is doing a lot of work hidden from view
This is not surprising. Historically, successful services or products provided via computer applications for mass use have been user-friendly enough to make consumers forget that a computer is doing a lot of work hidden from view. Buying with a credit card is one example.
The success of the smartphone is another example. The Blackberry phone introduced in the 2000s had many useful capabilities. However, accessing these services was cumbersome, involving a large number of steps.
The introduction of the Apple phone was revolutionary in that the access to services was through a touchscreen, eliminating the need for clicks. The built-in software and computing capability were cleverly designed to do all the hard work for consumers.
The modern car is another example of hiding the hard work from consumers. Starting a car electronically rather than through an external hand-crank was one of the early innovations that made driving accessible to a large consumer population. Features were added to the car that allowed people to get in their vehicle and start driving without having any knowledge of mechanics or electronics.
People are not reluctant to access computer data to assist driving. However, consider a situation where there is no computer-inputted address for a destination, one that is known only by its appearance and general location. How do you program the driving computer?
While people have adjusted very well to the idea of putting in a well-defined destination and getting directions through applications like Waxe, they remain in control of the car – the driver can adjust at their whim and sometimes ignore the directions that are provided. Hence, individual freedom of movement – the big incentive for driving a car and not taking the bus – is preserved.
So where does this discussion lead? Of course, cars and trucks can be electronically programmed to operate in environments where the path and traffic are relatively predictable. Think of a shuttle service in a campus environment or a dump truck in a controlled-traffic mine environment.
However, general consumer acceptance of unlimited self-driving technology is a different matter because so many conditions can arise to challenge a driver and undermine traffic safety.
Improving technology will allow more features to be added to cars to increase their safety and ease of use, such as sensors that detect hazards, but ultimately the everyday driver will continue to enjoy the experience of being in control of their car.
What is this author talking about? For one, self-driving cars aren’t in the "distant future"–they are already on the streets of the United States working out the last 1% (the hardest parts) before they launch on a large scale. Secondly, maybe in the "distant future" autonomous cars will be linked to one another via a wireless network, but right now, most AV software does not require links to the outside world–they are self-contained systems of cameras and LIDAR, as well as GPS.
Comments are closed.