It’s Time to Consider the User Experience of Self-Driving Cars

Tamir Hassan
Jul 30, 2017 · 3 min read

There is much hype around autonomous vehicles, and many are expecting a revolution to happen in the way that we get from A to B in the near future. Gone is the expectation that the self-driving car simply “slots into” the current ecosystem of roads, car parks, service stations, auto dealerships, etc. Instead, many people, including Lyft co-founder John Zimmer, are rightly beginning to question the whole concept of car ownership and why we allocate so much public space in our cities to car parking.

The potential of such a revolution is enormous, enabling us to regain street space used for parking and reducing traffic congestion due to carpooling being an integral part of the system. But in order for this to work, vehicles need to be fully autonomous, able to travel unattended to their next customer, pull up at the kerbside, take them to their destination, let them off, and drive on unattended to their next customer (or to an out-of-town parking area), and so on.

Despite advances in AI, this dream is still far away from reality. Whereas motorway driving is relatively simple to master, driving in city streets is a different game altogether; one which poses many yet unsolved problems for a machine. In particular, pedestrians and cyclists seem to be the most vulnerable when coming into contact with this technology, as they are much more difficult to detect than cars.

Whereas much attention has been paid to the technology built into the car for recognizing obstacles, I see this problem as having just as much to do with the design of the road system itself and how autonomous (and non-autonomous) vehicles interact with humans. This is an area where better design can mitigate the limitations of the technology, leading to a safe environment for all.

Case in point: autonomous rail systems, of which there are several in operation today. This is a no-brainer, as it’s a closed system: the location of each vehicle is known and, as with heavy rail in general, trains have absolute right of way; there is no expectation of people or other obstacles on the rails. But for light rail networks, such as trams, this is not the case. In many cities, particularly in Europe, trams share the road with other vehicles and even run across pedestrian areas. And even though they have a longer stopping distance than a car at a given speed, their interaction with other road users works well, as the track provides an important visual cue to us all, helping us anticipate their movement.

I am not suggesting that we should build a network of rails for self-driving vehicles, which is costly to maintain and would only be practical in cities. But this example gives us a good opportunity to rethink the way roads in cities are built in order to facilitate interaction between autonomous vehicles and other road users.

One option would be to paint “virtual tracks” on the road that the autonomous vehicle would stick to, veering off this path only when turning into another road or pulling up at the kerbside (possibly being guided by additional markings). In contrast to rails, paint is cheap and provides other road users with an important visual cue. Coupled with the requirement of fully autonomous vehicles to flash a light or otherwise be easily distinguishable from other traffic, and sensible speed limits for urban areas, this should lead to a much safer situation for all road users.

Tram tracks running through the main square of Brno, Czech Republic. The entire square is pedestrianized, yet both trams and pedestrians manage to co-exist in the same space

Off the city streets and on the motorway, we now have an environment that more closely resembles heavy rail, which is far more controlled and there is no expectation of pedestrians, cyclists or other obstacles. As recognition technology will never be perfect, I can imagine a future where every vehicle, self-driven or not, will be required to carry a virtual “beacon” that communicates with other self-driven cars in the vicinity, signalling its whereabouts. In this situation, a self-driving vehicle is arguably safer than one driven by a human, as the reaction time is greatly reduced.

Researcher in the Document Engineering field, musing about technology and how we interact with it

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade