Fitting autonomous vehicles with robotic, artificial eyes could make them safer for pedestrians.

That’s the surprising conclusion from a new study conducted by a research team from Japan’s University of Tokyo.

Explaining the purpose of the research, Professor Takeo Igarashi from the Graduate School of Information Science and Technology, said: “There is not enough investigation into the interaction between self-driving cars and the people around them, such as pedestrians. 

“So, we need more investigation and effort into such interaction to bring safety and assurance to society regarding self-driving cars.” 

The study’s 18 participants (nine male, nine female) had to decide whether to cross a road in front of a moving vehicle or not. For obvious safety reasons, these tests were done in virtual reality.

The vehicle in question was a self-driving golf cart that was fitted with two large, remote-controlled robotic eyes, and labeled the “gazing car.” 

Four scenarios assessed if people would cross the road in front of the moving vehicle – two where the cart had eyes and two without. The cart either noticed the pedestrian and was intending to stop, or failed to notice them and was going to keep driving. When the cart had eyes, they would be looking toward the pedestrian (when it was going to stop) or looking away (when it was not going to stop).

The participants experienced the scenarios multiple times in random order and were given three seconds on each occasion to decide whether or not to cross in front of the cart. The researchers recorded how often they chose to stop when they could have crossed, and how often they crossed when they should have waited.

The study had some surprising results, including a marked difference in behavior between the sexes. In particular, it found the male participants often chose to cross when the car was not stopping, but these errors were reduced by the cart’s eye gaze. 

In contrast, the female participants were found to make inefficient decisions rather than dangerous ones, such as choosing not to cross when the car was intending to stop. Again, these errors were reduced by the cart’s eye gaze. 

Despite the different results, what was clear was that the experiment showed that fitting eyes resulted in a smoother or safer crossing for everyone.

The general idea is, of course, that the eyes communicate the intent of the car.

While acknowledging the limited scope of the study, the researchers feel the idea is worthy of further investigation. 

“If eyes can actually contribute to safety and reduce traffic accidents, we should seriously consider adding them,” said Igarashi. “I hope this research encourages other groups to try similar ideas.”