ISU, UI Explore New Avenues in Traffic Research
by Aaron MackTweet
MATC faculty from Region VII’s University of Iowa (UI) and Iowa State University (ISU) are collaborating to combine two powerful software-based tools for transportation research. The project will link miniSim™—a human-operated desktop driving simulator running the same software as UI’s National Advanced Driving Simulator (NADS)—with the traffic microsimulation software VISSIM, a program that generates simulated traffic streams using researcher-controlled parameters.
Because it uses human operators in a simulated environment, miniSim™ is an ideal tool for collecting realistic data on driver behavior in the lab; however, as noted by Dr. Susan Chrysler, Director of Research at NADS, driving simulators are somewhat limited in their ability to accurately represent the behavior of ambient traffic. This inhibits the accuracy of simulator-based studies in terms of representing the driving environment.
“Built into [driving simulator and microsimulation] software,” explained Chrysler, are assumptions about how other vehicles will behave, such as how much headway they’ll leave between themselves and other vehicles, or how hard they’ll accelerate and decelerate. But these assumptions aren’t necessarily based on data, and are somewhat simplistic.”
Accurate input parameters such as speed and acceleration profiles are important for simulation-based traffic studies. Mobile source emission models, for instance, rely upon accurate estimates of speed and acceleration. The output from simulation models based on simplistic algorithms could lead to highly inaccurate estimates of emissions.
One of the strengths of VISSIM is that it allows researchers to program an ambient traffic stream to follow a number of different parameters; for example, traffic flow in VISSIM can be programmed to exhibit high volume or aggressive acceleration patterns; these parameters can be derived from statistical field data, then calibrated to a high degree of accuracy. Integrating this capability of VISSIM into the miniSim™ software would enable traffic researchers to simulate ambient traffic with a high degree of control and accuracy within human-operated driving simulations.
Conversely, linking miniSim™ and VISSIM would enable researchers to input parameters into VISSIM microsimulations that would be derived from human operators. VISSIM’s parameters are typically based on theoretical assumptions about how drivers will behave; the miniSim™ simulator could be integrated in order to provide data based on actual driver behavior obtained from subjects operating the simulator.
Dr. Shauna Hallmark, Interim Director of ISU’s Institute of Transportation (InTrans), noted the potential usefulness of supplementing traffic microsimulation with data from a human-operated simulator:
“If you took a person who was actually driving and you fed that information back into a microsimulation model, you could actually create more realistic speed and acceleration profiles; for example, whether drivers brake really hard or whether they brake gradually, things like that. You could actually make the microsimulation model more realistic because you add real world driving,” Hallmark said.
Both researchers noted that the potential benefits of the integrated system for traffic analysis were numerous. One specific application planned for the near future is the modeling of vehicle emissions and air quality impacts based on driver throttle usage—which would benefit equally from the accurate traffic representation and naturalistic driver behavior data provided by the integrated, “human in the loop” system.
According to Chrysler and Hallmark, integrating the VISSIM software with miniSim™ has been very successful. The next, somewhat more complicated phase will involve enabling the sharing of data about multiple vehicles in both modeling environments.
“That could to prove to be more challenging,” Hallmark said.
As noted by Hallmark and Chrysler, the ongoing project has provided a valuable opportunity for the two Region VII universities to collaborate toward a mutually beneficial research goal. The researchers also wished to acknowledge PTV America for providing VISSIM technical support.