Autonomous vehicles get ‘x-ray’ vision

Researchers in Australia claim to have developed a technology that gives autonomous vehicles ‘x-ray vision’ to detect pedestrians and cyclists in blind spots.

The technology is said to allow autonomous vehicles to track running pedestrians hidden behind buildings and cyclists obscured by larger cars, trucks and buses.

Funded by the iMOVE Cooperative Research Centre, the three-year project involved collaboration between the University of Sydney’s Australian Centre for Field Robotics and Australian connected vehicle solutions company Cohda Wireless. 

Cohda is commercialising the technology’s applications, which involve an emerging technology for intelligent transportation systems (ITS) called cooperative or collective perception (CP).

Using ‘ITS stations’ — roadside ITS information-sharing units equipped with additional sensors such as cameras and Lidar — vehicles can share what they ‘see’ using vehicle-to-X (V2X) communication. Researchers believe that the technology will benefit all vehicles, not just those connected to the system.

Leeds aims to improve safety in autonomous vehicles

Bad weather dataset could aid autonomous vehicles

“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,” said Professor Eduardo Nebot from the Australian Centre for Field Robotics.

Nebot explained that a connected vehicle was able to track a pedestrian visually obstructed by a building with CP information ‘seconds before’ its local perception sensors or the driver could possibly see the same pedestrian. This could provide extra time for the driver or navigation stack to react to the hazard, he said.

Another experiment showed the CP technology’s ability to safely interact with walking pedestrians, responding based on the perception information provided by the roadside ITS station. The expected behaviour of a connected vehicle when interacting with a pedestrian rushing toward a designated crossing area was also demonstrated.

“Using the ITS system, the connected autonomous vehicle managed to take preemptive action: braking and stopping before the pedestrian crossing area based on the predicted movement of the pedestrian,” said Prof. Nebot.

autonomous vehicles
CP-enabled vehicle detects a vehicle obscured by building. Image: Cohda Wireless

“The pedestrian tracking, prediction, path planning and decision making were based on the perception information received from the ITS roadside stations.”

Cohda Wireless chief technical officer Professor Paul Alexander said that the new technology breaks the physical and practical limitations of onboard perception sensors, and ‘embraces improved perception quality and robustness’.

“This could lower per vehicle cost to facilitate the massive deployment of CAV technology,” Alexander said.

He added that using CP for manually driven connected vehicles also brings ‘an attractive advantage of enabling perception capability without retrofitting the vehicle with perception sensors and the associated processing unit’.