Collective perception research could give cars the ability to see through obstacles

A limit of self-driving cars and humans alike, is that they can only see what they can see.

Eyeballs, cameras, lidar and radar are all line-of-sight technologies, so if something is obscuring the signal or light a brain or computer has no idea what's going on behind it.

But a new technology called collective perception could give vehicles an 'x-ray' sense to see what they currently cannot.

4

The University of Sydney's Australian Centre for Field Robotics, in collaboration with the iMOVE Cooperative Research Centre and Aussie software supplier Cohda Wireless, has released new research which combines data gathered by "intelligent transportation system stations" (ITS-S) with others on the road using vehicle-to-X (V2X) communication.

According to the findings, this allows vehicles to find out what’s going on around them – even if their direct sensors are obscured.

Sounds fancy – but think of ITS-S as sensors either on a car or on the street (cameras or lidar for example) and V2X communication as a network cars can access to get information. It's basically using other cars as your eyes on the road in real time.

4

"This is a game changer for both human-operated and autonomous vehicles, which we hope will substantially improve the efficiency and safety of road transportation," says Professor Eduardo Nebot from the Australian Centre for Field Robotics.

"[In the research] the connected vehicle was able to track a pedestrian visually obstructed by a building. This was achieved seconds before its local perception sensors or the driver could possibly have seen the same pedestrian around the corner – providing extra time for the driver or the navigation stack to react to this safety hazard".

An example of collective perception in action could be detection of a cyclist riding alongside a bus that would be obscured to one vehicle, but visible to another travelling in the other direction.

4

The cars can share what they see – so if the cyclist makes a sudden move, the car that couldn't previously see the cyclist is now aware of them and can take evasive action if required. Such a situation is impossible with current sensor technology.

Cohda Wireless' Chief Technical Officer, Professor Paul Alexander, says "collective perception enables the smart vehicles to break the physical and practical limitations of onboard sensors, and embrace improved quality and robustness".

It is hoped that by tapping into the sensors of all the vehicles on the road, the depth of knowledge of all cars improves and can provide a safer experience not only for motorists, but for pedestrians and cyclists too.

There's no firm timeline for when collective perception technologies will hit the road, and the standards for vehicle-to-vehicle communication are still nascent, but the research is promising and could deliver an autonomous driving experience that is not only as good as a human driver, but superior.