Alicja Halbryt

Oct 8, 2019

4 min read

Make us, pedestrians, aware.

The future of self-driving cars.

The cover of The Times features an article about the first ever protest of Londoners against self-driving cars. They feel their freedom and rights are being taken away. People accuse the government and car manufacturers of putting tech over people. They claim they were not prepared. Passengers protest too — they are frustrated with pedestrians bursting into the street and disturbing the flow of their journey. An older lady says she doesn’t want to die hit by a car. She overheard that it is more likely to happen to her than to younger people.

Wait. What?

The above article comes from 2025 and is a future artefact created for the purposes of my Master final project. The thesis explores the future of pedestrians among self-driving cars. It aims to design ways to prepare London pedestrians for daily interactions with the vehicles during the transition period — a time when there is a mix of self-driving and non-self-driving cars on the streets. Which might happen within the next 5–7 years, or less.

The ethics around self-driving cars is an endless journey through fascinating, controversial, and often unsolvable issues. In fact, I came across one while looking for a good topic for the Master thesis. It was an article saying that people of darker skin tone are more likely to be hit by a self-driving car, which is a result of biased algorithms. The cars are simply better at recognising white faces. Fortunately, the issue has a potential to be resolved in the future, in contrary to so many other nuances. Like the ‘driverless dilemma’ — the on-going discussion over who should the self-driving car kill or save in an unavoidable accident (Moral Machine by MIT Media Lab). Should it be a child or an older lady. A passenger or a dog. 3 pedestrians or 1 passenger. A professor or a homeless person. Even though situations like this will probably hardly ever happen (the AVs are supposed to be extremely safe after all), and the cars won’t really know the person’s social status, they will need to know what to do. Namely, they will need to be programmed to know what to do. Who to save. So, who will dare to program someone’s death? And if someone dies, whose fault will it be? Maybe cars will be set to always save the passenger? How would this impact the pedestrians? And if that was the case, how should people behave on the street in order not to get killed or harmed?

Examples above show that pedestrians are considered an obstacle rather than part of the traffic and the city. Cities are in fact built for cars. The roads facilitate getting from point A to point B in a vehicle. Pedestrians were given narrow pavements and crossings. Some avoid these and choose to jaywalk instead. This poses a major challenge for self-driving car makers. Not only because the cars will need to be able to stop in time to avoid a collision, but also because the cars’ safety level might be so high that pedestrians will use it to their benefit. They might disturb the AV flow, knowing that it will certainly stop. It could result in the car breaking every couple of meters. Yes, it will annoy the passengers.

So, on one hand there is a threat of getting killed by a piece of algorithm set to hit some group of people. And on the other hand, there is a likelihood of the cars being so safe that it might become unbearable for the passengers. If I was more likely than others to be hit by a self-driving car, I would like to be made aware. If my car stopped every time a person appears a few meters ahead of it, I would like that person to be made aware.

In a survey conducted for my Master project, I asked people about their views on future pedestrian interactions with autonomous vehicles. I asked if they were willing to adapt new behaviours in order to stay safe on the street (e.g. raise their hand to inform a car of their intention to cross). Another question investigated whether people would like to be educated on the DOs and DONTs around self-driving cars. And how they would react if they knew self-driving cars were more likely to hit their demographic group. From over 80 responses I found that the majority of respondents (75%) would change their behaviours to maintain safety. 90% would like to know the DOs and DONTs. And they would feel scared, avoid the streets, and even protest against self-driving cars. These responses inspired the future article above.

Is rising awareness the answer to the future problems the society is likely to encounter? How can people prepare for the coming of the technology which is believed by many to be the future? How might we make it a future which people know well instead of fear?

Answers coming in December.