A glance, a nod, a hand signal: much of what makes up communication and interactions between people on the road is impossible when the other vehicle is a self-driving car. From LEDs to light projections or emojis – different approaches in the industry all have the same goal: getting cars to talk.
Sometimes you don’t know what you have until it’s gone. This may have been how some pedestrians in downtown Cologne felt a few weeks ago when they came across a car with something special about it: it lacked a driver. At least a clearly recognizable one – the driver had been disguised for a video shoot. Together with Chemnitz Technical University, Ford developers were testing people’s reactions to cars that seemed to be driving autonomously.
Ford is not alone in this: most automobile manufacturers are investing a lot of effort in studying self-driving cars. The key difference from previous developments is that – depending on the technology available – self-driving cars transport passengers, not drivers. In other words, the more decisions formerly made by human drivers that the vehicle makes, the more “human” it has to become.
After all, traffic is always a form of interaction and communication. This involves hand signals and communication by nodding, gestures, and glances. For the future of mobility, this means that self-driving cars have to learn to correctly interpret signals coming from human road users. On the other hand, people nearby also need to know what a self-driving vehicle intends to do next. If such communication between man and machine is to work, the developers at automotive companies need to make it possible for cars to talk. And how is this done?
Communication using colors
The Ford test car relies on different color signals. Using illuminated panels on the roof, it shows what it will do next. In a video about the experiment that Ford posted, viewers can see how people react. First, they look at the driver. Surprised, they then notice the color signals sent by the car: green for “go ahead”, red for “stop”. “Uniform standards for color signals in all car brands would certainly be a good idea,” says Monika Wagener from the Ford Research and Innovation Center in Aachen.
Communication with signals is better the more intuitive they are, so they should be familiar to all road users. There are various ideas on how to do this: from smileys on the car roof to LEDs in the grill, to projections such as pedestrian walkways or arrows showing the intended direction of travel – there are many possibilities on the road towards the mobility of the future. After all, one thing is clear: fully autonomous vehicles are not an innovation that will be a common sight in German cities within the next two to three years.
However, level 2 semi-autonomous vehicles may already be on German roads in a few years. In these vehicles, intelligent systems support human drivers. It is likely that they will start off on highways, commuter routes, and other roads with relatively clear structures.
Daimler has called its research vehicle a “cooperative vehicle”. The Stuttgart-based company has also chosen lights to improve communication and understanding in traffic. A recent study concluded that other road users want autonomous cars to be marked as clearly and extensively as possible, as a company spokeswoman says: “People have to immediately and intuitively recognize what an autonomous vehicle is about to do.” An example of this would be slowly blinking lights on the car roof signaling that the car is braking. On the other hand, quickly flashing lights would indicate that the car is about to start moving.
Before driverless vehicles can go into mass production, both technical and legislative obstacles have to be overcome. One of the biggest challenges in this regard is programming an autonomous car so that it does just what human drivers do again and again in certain traffic situations because they have to: breaking traffic rules. For example, driving over a solid line on the road to create enough space for an ambulance to pass. This also raises the question of how an autonomous car would inform nearby road participants of what it intends to do in such a scenario. And of course there is another challenge to be overcome, one which the Ford developers encountered during the company’s test drive in downtown Cologne: “Some people did not even notice that our test car did not have a visible driver,” says Monika Wagener. “Because they weren’t paying close attention to who else was using the road.”