The march of driverless technology continues apace, and whilst it’s relatively straightforward to see how car-to-car communication will occur, it’s much less clearcut just how well driverless cars will communicate with humans, whether on the roads themselves or as pedestrians.
I wrote earlier this year about a study from researchers at University of California, San Diego that produced an algorithm that can detect a pedestrian in near real-time and with high levels of accuracy.
“We’re aiming to build computer vision systems that will help computers better understand the world around them,” the authors said.
A number of projects are underway to develop a greater understanding of how we behave, and to thus build a predictive element into the vehicles programming.
These improvements are significant, but there will still be a requirement to overtly communicate with pedestrians and other road users if driverless cars are to be safe.
That’s the aim of a new startup called Drive.ai that was formed by a group of former Stanford researchers to try and ensure driverless cars better understand human behaviors.
To do this, they’re equipping test vehicles with a range of display and sound systems to allow the car to communicate with pedestrians. The aim is to help the company gain a better insight into just how humans behave when interacting with vehicles on the road.
The emergence of driverless technology are going to fundamentally change how we behave around road vehicles, and as such the behavior of the vehicles themselves is likely to be different to today. It will inevitably take a while for this change to unfold and a best practice to emerge, but the team hope that by considering it from the outset they will be better equipped from day one.
Managing the transition
As is increasingly common, the company’s first product is designed to modify an existing vehicle so that it becomes driverless. The aim is to empower company’s with a fleet of vehicles to go driverless. Each vehicle will be fitted with the standard array of sensors to control the vehicle, but will also come a communication system fitted to the roof.
Through this, a range of messages, sounds and lights will be used to communicate both with pedestrians and fellow drivers. The system might say, for instance, that it’s safe for a pedestrian to cross, or that it’s giving way to another motorist. It’s designed using deep learning and the team hope that it will be a valuable step on the road to automation.