Googly Eyes On Autonomous Vehicles Helps Pedestrians To Decide On Crossing Roads

Even if researchers are able to perfect autonomous vehicles, there is the interaction between them and pedestrians to consider. With there being no driver and only passengers in the vehicle, it is impossible for pedestrians to gauge whether the vehicle has sensed their presence as there no longer an indication from the driver. Hesitation could very well be the cause of accident as the pedestrian deliberates “to cross or not to cross”.

Japanese researchers have come up with a simple and quite amusing solution, that is to fit robotic eyes on the autonomous vehicles. The robotic eyes would look at the pedestrian to acknowledge their presence or away if it has not been registered.

The team from the University of Tokyo tested the scenario with VR by modeling a golf cart fitted with “Pixar eyes” and four scenarios, two with the eyes and two without. 18 participants underwent multiple tests in random order, during which they had three seconds to decide whether to cross the road. The researchers measured the error rate, ie. they chose to stop when they should have crossed and vice versa.

In the VR simulations, they found that overall, pedestrians made “safer or more efficient choices” when the cart had the eyes. Interestingly, in the scenarios without the eyes, the male participants were more likely to cross the road when the car was not stopping and female participants were more likely to not cross when the car was intending to stop. These two errors were reduced when the cart had eyes.

Of course 18 is a very small sample size. “We need more investigation and effort into such interaction to bring safety and assurance to society regarding self-driving cars,” said lead author, Professor Takeo Igarashi.

Nonetheless, it’s an interesting idea that could contribute to future road safety. If not googly eyes (although, we quite like them), then some sort of external indicator to show that the driverless car “sees” the pedestrian or is about to slow down or speed up would contribute greatly to future road safety.