A state of the art driving simulator combined with computerized traffic scenarios is helping U of U researchers better predict traffic behavior.
A super-computer backed driving simulator at the University of Utah Traffic Lab is helping researchers see how traffic really behaves. “We can provide the state of the art,” says Devin Heaps, Administrative Manager at the Utah Traffic Lab. Heaps is confident that the lab can help researchers “push science forward” when it comes to the study of how to better manage traffic.
The U’s simulator is “the first of its kind to marry an individual physical person driving with a driving simulator,” according to Heaps. The simulator was originally developed to train soldiers on how to drive military vehicles through the streets of Bagdad. The traffic lab takes the system beyond driving practice and allows engineers to collect data on how individual drivers react to realistic computer generated scenarios.
Measuring the actions of individual drivers “builds a highly realistic picture of actual traffic flow,” says Professor Peter T. Martin, Director of the Traffic Lab. Traffic engineers commonly use macro simulation, which measures traffic flow and pressures in platoons of cars. Measuring in units, or micro simulation, is the key to getting a more realistic picture of traffic flow that can lead to better real world solutions.
The simple simulator made news a few years ago when psychology researcher David Strayer found that using a cell phone while driving is like being legally intoxicated. Today, the traffic lab is using data collected while subjects drive in realistic roadway scenarios to show how cell phone use impacts a roadway or an entire transportation network.
Researchers have also demonstrated some engineering solutions that make for a safer roadway. Wide roads with big signs and very few curves make drivers comfortable – and give them a bit of a lead foot on the gas pedal as well. Using micro simulation, researchers can show that narrower roads with curves can be safer because drivers are more cautious and not so inclined to speed. Researchers can also model upcoming transportation projects or new road features. “We are planning to model Flex Lane use,” says Martin.
Micro simulation can also be used to predict driver behavior during a crash or other incident that blocks lanes on a freeway. Scenarios that reflect actual or potential incidents can be programmed into the simulator. Then, researchers can collect data from individual drivers and predict delay on the roadway and the rest of the network. The precise data can guide UDOT TOC operators, who control signals and overhead signs, to make better real-world decisions about how to manage traffic during actual incident.
“These tools enable traffic engineers to answer ‘what if’ questions about traffic behavior. Millions of calculations and comparison to actual flow measurements enables these simulations to model traffic very realistically,” explains Martin.