A study by engineers from the University of New South Wales proposes a freeway network design with exclusive lanes for autonomous vehicles (AVs).
Using computer modelling of mixed scenarios, they found dedicated lanes significantly improved the overall safety and traffic flow in a hybrid network of pedestrians, cyclists, automated vehicles and legacy vehicles, the study published in the Journal of Transportation Research Part C: Emerging Technologies.
If the road and transport network is not prepared for these vehicles when they enter the market, it will significantly hinder the travel experience of all road users, lead author Dr Shantanu Chakraborty from UNSW School of Civil and Environmental Engineering says.
“Traffic congestion costs the economy billions of dollars every year in all the extra time spent commuting. The proposed model will help minimise interaction with legacy vehicles and reduce overall congestion on the road,” he says.
“The mix of autonomous vehicles and legacy vehicles will cause issues on the road network unless there is proper modelling during this transition phase. If we get caught out and we’re not ready, we won’t reap the full benefits of the technology behind these autonomous vehicles.”
Adding an exclusive lane for autonomous vehicles means removing a lane from legacy drivers – so this may cause a little disruption, he says.
“If you look at our existing network, we already have something similar with dedicated bus lanes – so we’re not reinventing the wheel here.
“Freeways are also the best network of car lanes to trial as they have dedicated entry and exit points where drivers can automatically switch on and off their automated features.”
AVs not only have the potential to provide cost-effective mobility options, but road users can reap the benefits of reduced congestion.
Chakraborty says road users can activate the autopilot features of their vehicles while they are in these exclusive lanes. The automation of movement of the vehicles means the flow of traffic would significantly improve in these lanes as drivers are not solely relying on their attention and reaction time to traffic conditions.
“Say you’re sitting in traffic and the traffic light turns green, the driver doesn’t instantaneously take off that second; there is usually a response time before you press on the pedal and the car moves,” he says.
“Then the driver behind you reacts and so forth and by this stage, there has been some time passed.
“However, with autonomous vehicles, the movement is more coordinated because the vehicles are fitted with sensors. When the signal turns green, all the vehicles move simultaneously which will improve traffic flow and reduce congestion.”
Chakraborty says variable signboards could be used to change the lane designation based on the traffic condition at the time. This will mean during peak hours roads can be used more efficiently depending on the traffic conditions at the time.
“Our modelling accounts for changing traffic conditions. For example, during non-peak hour times when we don’t need a lane for autonomous vehicles, we can have all lanes open for legacy vehicles,” he says.
“Due to the minimal infrastructure, our proposed model also has the potential to design ramp metering for freeway networks to help regulate the flow of traffic during peak hours.”
Similar to existing high-occupancy lanes, transit lanes or T2 or T3 lanes, a fine can be applied when drivers of legacy vehicles enter AV dedicated lanes, Chakraborty says.
“Like any other road rules, we can only trust that drivers obey the signs and road rules.”
Chakraborty won the Aspire Award in UNSW’s Three Minute Thesis competition last year, presenting his work on lanes for self-driving vehicles.
