In 2019 pedestrian fatalities in the U.S. reached the highest in 30 years and represented a near 50 percent increase over the past decade. Attempting to arrest this trend is “Vision Zero” – a strategy to eliminate all traffic fatalities and severe injuries, while increasing safe, healthy, and equitable mobility for all. Nevada is one of several states in the U.S. that are part of this initiative. Against this context, a company called Velodyne LiDAR has partnered with the University of Nevada and the state to install sensors around the city of Reno and other areas in Nevada. The program aims to turn Nevada streets and roads into a lab, with the end goal being to integrate LiDAR sensors into the built environment and provide constant data and feedback to a network of connected autonomous vehicles.
The data that LiDAR systems are able to collect and transmit may prove valuable for continuing to refine the algorithms that power self-driving vehicles and smart cities. The trajectory data, especially at intersections, is what self-driving algorithms need to function safely and reach their maximum potential. According to Associate Professor Hao Xu from the University of Nevada, autonomous vehicles are not entirely fail-proof at this point. They still struggle to predict and react to sudden movements at corners. Through the use of these additional sensors embedded into the urban infrastructure, some of the glaring holes in the current dataset that fuels self-driving algorithms can be filled.
On the path to mass-production of fully autonomous vehicles, cooperative perception, which is the exchange of local sensing information with other vehicles or infrastructures via wireless communications, roadside LiDAR systems mounted to existing infrastructure like traffic signals, bridges and buildings, could be the way forward. The perception range can be considerably extended up to the boundary of connected vehicles. Other perception sensors, systems, and modalities could of course also be considered.
Another interesting approach could be to use AI to generate simulated rare traffic situations from collected real rare situations, for example near-crash events. The situations generated could potentially be used to “teach” the autonomous vehicles how to react rapidly to avoid a crash in unusual settings.
Vision Zero could certainly be possible through the implementation of autonomous vehicles and increased investment in smart infrastructure like roadside LiDAR systems powering. Even though this technology is still in its infancy, a truly synchronous and connected transportation network is within reach.
Written by Joakim Rosell,
RISE Mobility & Systems (Människa-autonomi)
Do you have any comments of your own about this article? Please write to us.