Self-driving cars; what’s the issue?

I find the Google car very interesting (now under the business name “Waymo”), as well as the self-drive mode of Tesla and I believe Toyota cars. Uber is trying out their own cars too. There are plenty of issues to discuss on this topic.

Just imagine how great this would be for so many people:

  • Elderly can get out and about
  • Handicapped, for example wheel-chair users who are not comfortable driving, can easily be transported
  • Blind people can be more independent
  • No more drunken drivers (see text further down)
  • No more drugs or medicine-affected driving (see same text as drunk drivers)
  • Children can be taken to school “automatically”
  • A significant, extreme, drop in traffic accidents
  • You can focus on other things than driving during your commute

In fact, The Oatmeal summed it up pretty nicely in his comic after driving a Google car: 6 things he learned from that ride.

Here’s a video showing the Google car and why a blind man really wants this kind of technology as soon as possible. As the golden-aged blind man says; he can finally be able to go somewhere without making arrangements with others for driving him, he can decide himself and be spontaneous. It gives him great freedom:

Ethical dilemmas
Regarding the “crash for the greater good”, Jean­ Francois Bonnefon at the Toulouse School of Economics in France decided to find out what the public thought of self­-driving cars and ethical dilemmas. The “crash for the greater good” scenario is this:  what if there is a woman in the car, maybe even pregnant (so, two lives), realising she is driving at high speed towards a crowd. There is a mountain wall on one side of her and a steep drop on the other side. Should she crash her vehicle into a mountain wall knowing she might not survive (for the greater good), or should she risk killing several people in the crowd?

The researchers asked several hundred employees at Amazon.com, and the result I found the most interesting was this: “And therein lies the paradox. People are in favour of cars that sacrifice the occupant to save other lives—as long they don’t have to drive one themselves. (MIT Technology Review, 2015)

In other words, it sounds like typical human psychology: we all agree on something, as long as it doesn’t affect us personally.

However, many people argue that this is a false problem. How often can it be said that this scenario has happened, where the driver was clear­headed, and not having a heart­attack, that is? Which, the self­-driving car won’t be of course. In other words, the car would see this crowd and slow down to a stop, which they often do when confronted with an uncertain situation. (Knight, 2015) The only scenarios that might be possible could be blown tire or skidding on ice. (DarrelR, Oct 28, 2015, comments section on (MIT Technology Review, 2015))

Are we overthinking this? Is it really a good argument against a self­-driving car; the worst possible scenario that probably won’t happen? Maybe we are being paranoid?

Maybe not, though. A more common scenario could in fact be that a child suddenly runs out in front of the car, chasing a football. Should the car swerve into that oncoming van or hurt the child? Would it be better to hurt the people in the car – maybe children are in the car? These questions could be quite difficult to program an answer for. (Knight, 2015)

In any case, ethical questions and engineering are starting to go hand­ in­ hand. Stanford University recently gathered engineers and philosophers to discuss ethical dilemmas with self­-driving cars. (Knight, 2015)

The introduction of self­-driving cars will definitely make for a safer traffic every day. Considering the high numbers of fatal traffic accidents, some even believe it unethical waiting to introduce these cars just to get a clear philosophical answer for the topics above. (Knight, 2015) Can we afford to wait for an ethical answer for this, knowing we could have saved many lives while we are discussing?

What about maps, for the routes?
In the hills close to my house, every year,­ every single year, ­I read in the newspapers that some truck has skidded off the road. Every year, like clockwork when the snow has come. Not only is it because they lack winter chains on their tyres but also because many have put the shortest GPS route instead of the fastest. So, bad planning on the human part!

In the case of a traffic accident or road closures ­ how rapidly will the GPS network be updated? Today, it’s not as quick as we would like it to even though there’s many volunteer updaters out there. There’s different systems though so different efforts are not joined ­ MapQuest, Google maps with GPS, TomTom, Garmin etc. Lots of volunteer updating, but not joined in 1 common place that I’m aware of. The cars should try to join forces to get the most updated GPS maps.

Where should human interaction be limited?
So, we would need to be able to tell the car to change direction, while we’re driving for example. This much should be allowed. But­ in regards to drunken drivers etc. ­ maybe this is all we should be allowed to do? Maybe there could be a red emergency stop button which would then stop the car: the car could then ask us if it should take another route to the destination or return back. Do we need much more?

I can envision different types of cars, maybe sold for different uses. Maybe you had to be over 18 to get a car where you could steer some. Maybe if below 18 years old, you could still get a car but you wouldn’t be allowed to overrule anything (except for emergency stop and route planning.) Could a parent put their children in a certain type of car that will take them to school? No ability to re­program once she/he has put the destination in, with a particular entry code? No option to drive outside of a certain area?

Then again  I guess the car would have to be strict too? “No, I won’t drive into the water. Don’t put that as a destination, please”?

Digital security
On another safety aspect; we need to consider hacking. British Telecom (BT) is now providing “ethical hacking” services to vehicle manufacturers, insurance companies and all other businesses in the automotive industry. (Larkin, 2015, pp.8­8) As we have more and more intelligent systems in our car, from 3G to WiFi, we have more ways of being hacked. Not only remotely, but what if someone slid a CD with code into the system’s CD-player? Or inserted a USB-stick or cable into the car and uploaded/extracted data?

BT offers the industry to test all of these things before a vehicle is released as well as ongoing testing later on. We have already seen that a car was completely hijacked. (Greenberg, 2015) Fiat Chrysler has recalled 1.4 million vehicles after this revelation by Wired. (The Guardian, 2015) Therefore, I believe this kind of testing to be a real necessity.

Summarized, I can’t wait for the self-driving cars to become a reality on the roads everywhere.

 

Bibliography

Knight, W. (2015) Self‐Driving Cars Get a Code of Ethics | MIT Technology Review [Online] Available from: http://www.technologyreview.com/news/539731/how‐to‐help‐self‐driving‐cars‐ make‐ethical‐decisions/ (Accessed: 29.12.2016).

MIT Technology Review. (2015) Why Self‐Driving Cars Must Be Programmed to Kill | MIT Technology Review [Online] Available from: http://www.technologyreview.com/view/542626/why‐self‐driving‐cars‐must‐be‐programmed‐to‐ kill/ (Accessed: 29.12.2016).

Greenberg, A. (2015) Hackers Remotely Kill a Jeep on the Highway—With Me in It | WIRED [Online] Available from: http://www.wired.com/2015/07/hackers‐remotely‐kill‐jeep‐highway/ (Accessed: 29.12.2016).

Larkin, J. (2015) ‘BT Assure Ethical Hacking for Vehicles’, Automotive Industries, pp.8‐.

The Guardian. (2015) Fiat Chrysler recalls 1.4m vehicles in wake of Jeep hacking revela on | Business | The Guardian [Online] Available from: http://www.theguardian.com/business/2015/jul/24/fiat‐chrysler‐recall‐jeep‐hacking (Accessed: 29.12.2016).