UNIVERSITY PARK, Pa. — Penn State researchers aim to develop a driver-in-the-loop vehicle simulator tool in order to better understand the effects that Connected and Autonomous Vehicle (CAV) technology has on commuting behavior, thanks to a $60,000 Multidisciplinary Research Seed Grant.
According to the National Highway Traffic Safety Administration, vehicular injuries and deaths in the United States have spiked for the first time in nearly 50 years. CAVs are shown to be one of the more promising technologies to improve driver safety. They also have the potential to reduce traffic congestion, improve travel efficiency and convenience and expand mobility options. More advanced high-level CAV technology, which includes Levels 4 and 5 on a 5-point scale, is able to perform driving tasks with little or no intervention from drivers, allowing motorists to perform other tasks.
While many believe commuters may start to favor longer commute distances in exchange for lower-cost homes if they are able to be productive during their commute, a key missing factor in predicting whether this is true is better understanding the effects of CAV technology on drivers’ acceptance of longer travel time and changes in commuting behavior.
“We’re trying to see if people are willing to spend more time in the car if it’s autonomous because they can do other things,” said Ilgin Guler, assistant professor in civil engineering. “Given different types of driving — active, passive or no-driving — how long and far would people be willing to commute for?”
Guler, along with Sean Brennan, professor of mechanical engineering, and Yiqi Zhang, assistant professor of industrial engineering, plan to use the grant to develop a driver-in-the-loop simulator tool that allows drivers to experience dynamic traffic flow simulations. This will help the researchers understand the effects of CAV technology on commuting behavior.
In order to do this, the team will first integrate a microscopic traffic simulator with a driving simulator to recreate traffic conditions in State College, Pennsylvania. This will involve developing the necessary software that will allow the driving simulator to interact seamlessly with a microscopic traffic simulator. The driving simulator will allow a human to drive a traditional, semi-autonomous, or autonomous vehicle, and the exact behavior of the car — including location, speed and acceleration — will be provided to the microscopic simulator.
“We want to make sure the signals are up-to-date, the traffic conditions are up-to-date, and try to match the congestion levels that we observe in State College to the simulator itself,” said Guler. “We're trying to make everything as realistic as possible, so that when we do look at the impacts of connected autonomous vehicles, we're basing it off of a real scenario.”
Next, the researchers will examine the driving simulator validity. For this step, human experiments will be conducted to collect and compare driving performance data in the developed driving simulator and the real roadway. These tests will be used to generate baseline data for general driving behavior such as speed control, lateral control, and responses to traffic lights and signs.
Finally, the research team will conduct human experiments to investigate the impact of CAVs on driver behavior and drivers’ acceptance of CAV technologies. The experiment will include Level 4 and Level 5 CAV technology, some with internet access and some without internet access. The automation level and CAV’s connectivity will be combined to yield four experimental conditions, which will be investigated for the change in drivers’ acceptance of longer travel times.
“We're trying to get to that time sensitivity in how people's value of time will change with different automation types,” said Guler.
In the end, the team hopes to have a much better understanding of the impacts of CAV technology on changes in driver behaviors at a community level and, eventually, how city structure will change if people move further away from urban centers.
“The future of traffic is autonomous, you cannot ignore it,” Guler said. “This research will lay the ground work so that when these Level 4s and 5s are developed and released, we will have the data needed.”
Established in 2014, the Multidisciplinary Research Seed Grant program aims to help faculty attract high-impact multidisciplinary and center-level research funding from the state and federal government, industry or foundations. This one-year project will conclude in January 2020.