Minnesota has been a leader in using ramp metering—traffic signals on highway entrance ramps—to control when vehicles enter highways to keep traffic flowing. The state installed its first meters in 1969 on Interstate 35E at the entrance ramps from Maryland Avenue and Wheelock Parkway in St. Paul.
But what will ramp metering look like in a future that’s expected to have more automated or self-driving vehicles?
That’s the question researchers at the University of Minnesota wanted to answer. In a study sponsored by the Minnesota Department of Transportation (MnDOT) and led by CTS scholar Raphael Stern, associate professor with the Department of Civil, Environmental and Geo- Engineering, researchers focused on learning how different vehicle automation scenarios—from low-level systems such as adaptive cruise control (ACC) to fully automated vehicles (AVs)—may influence ramp metering strategies.
In addition, the researchers considered modifying existing ramp metering algorithms to make them more effective under future scenarios in which more AVs are expected on our roadways.
For the study, researchers used MnDOT’s ramp meter algorithm along with traffic flow data to simulate five on-ramp sites in the Twin Cities metro area and modeled how different levels of automation affected the performance of ramp metering. The modeling included vehicles with no automation (with the driver responsible for all aspects of operation), ACC-equipped vehicles, and two types of AVs (with and without inter-vehicle connectivity).
“Vehicle automation will change the way traffic behaves,” Stern says. For instance, vehicles with driver-assist technologies such as ACC can make traffic worse because their systems tend to brake more abruptly and leave large gaps between vehicles, he explained. ACC systems are already common in new cars.
On the other hand, fully automated and connected vehicles can follow more closely and respond more smoothly, improving traffic flow and reducing delays.
Findings showed that increased vehicle automation could result in underused capacity on the road, depending on the type of automation. Simulations also revealed that AVs offer opportunities to enhance traffic flow on ramp meters.
How will that scenario affect future planning for ramp metering? Researchers proposed two adjustments to MnDOT’s existing ramp metering algorithm, based on the makeup of traffic flow. These adjustments considered how different vehicles behave—like how closely they follow each other or how quickly they react. The changes resulted in much less ramp waiting time across all vehicle types.
“Ramp metering is an effective tool to manage traffic and improve travel times, but it has to be set correctly,” Stern says. “We need to be prepared with the right settings in a future with more autonomous vehicles on our roads.”
Garrett Schreiner, MnDOT freeway operations engineer and the project’s technical liaison, says the results of the simulations using the adjusted ramp metering algorithms are encouraging. “This demonstrates tools MnDOT has at its disposal to mitigate traffic flow issues that may arise due to integrating automated vehicle technology,” he added.
Stern notes that insights gained through this research underscore the importance of adapting ramp metering strategies to the shifting dynamics of vehicle automation.
“Research findings will help us better manage infrastructure in the future, resulting in reduced congestion and travel times for travelers, as well as cost savings through a reduced need for road expansion,” he says.
Continued research could focus on further refining MnDOT’s metering algorithms by testing them in real-word situations and creating ways to estimate traffic patterns in real time.
—Peter Raeker, contributing writer
