A James Cook University researcher says cars that 'see through each other's eyes' by sharing data with neighbouring vehicles will be the next wave of autonomous vehicles.
Tao Huang, Senior Lecturer, Electronic Systems and IoT Engineering at JCU, is the lead author of a new paper examining how autonomous vehicles can become safer. The paper was published in Proceedings of the IEEE, a flagship journal of the Institute of Electrical and Electronics Engineers and one of the most prestigious and highly cited journals in its field. His PhD student Xi Zhou, is recognised as co-first author.
Dr Huang said current autonomous cars rely solely on their own onboard sensors.
"This means that each vehicle operates independently. With limited sensing range and obstructions on the road, this can lead to blind spots and unreliable detection of objects."
Dr Huang said autonomous vehicles will become safer and smarter by sharing what they see with one another.
"The approach, called Vehicle-to-Everything Cooperative Perception (V2X CP), allows cars and roadside infrastructure to exchange sensor data in real time. This enables vehicles to detect hidden dangers - such as a pedestrian stepping out from behind a truck, that would otherwise remain invisible to a single car," said Dr Huang.
He said the new paper provides a comprehensive overview of the latest technical advances in the fast-growing field.
"We cover methods for extending a vehicle's awareness, reducing blind spots and improving decision-making in busy environments. We also review the major challenges, such as communication limits, data alignment and privacy concerns," said Dr Huang.
He said the paper showed clearly that cooperative perception is rapidly evolving and has significant potential.
"Combining advanced communication technologies, multimodal sensing, and AI-driven data processing gives us a pathway toward reliable, efficient and ethically sound autonomous driving and intelligent transportation systems. This is going to mean safer, more effective and highly capable autonomous mobility."
Link to paper here.
