New technologies are being added to vehicles at a growing rate to assist drivers and even fully take over the driving task in some situations. These technologies are generally camera-based and typically rely on pavement markings to maintain vehicle position and navigate the roadway. As drivers become more reliant on these systems, and for these systems to meet their potential safety benefits, the pavement marking infrastructure needs to be optimized to provide adequate roadway delineation. What is somewhat unknown is how different this optimized pavement marking system is from current practice and how different are the visibility needs of vehicle systems and the human eye. This project explores how various pavement marking configurations impact automated driving systems' ability to track the markings and maintain lane position. Evaluations take place in Texas and Minnesota on closed-course, open-road, and pavement marking test areas. Various camera-based systems are used to view the markings and generate feedback on the impact of the marking configurations and characteristics on the ability of the camera systems to track the markings. The research team analyzes the test results and provides recommendations to improve pavement markings to increase the function and reliability of camera-based pavement marking tracking driver assist features such as lane centering. The goal is to improve and maintain markings so the driver assist features can function reliably and yield safety improvements by reducing crashes, especially run-off-road crashes.
Various broken lane line configurations, marking length and gap between markings, exist across the United States. The normal width of pavement markings and the use of contrast markings with broken lane lines also varies across the country. This project explored how various pavement marking configurations impact observations from drivers during an open road human factors evaluation. The participants drove an instrumented vehicle through test areas with various pavement marking configurations. Participants provided feedback on marking visibility and their preference toward the pavement marking pattern. Participant observations took place during the day and at night. Researchers also conducted a preliminary investigation into the same marking test areas using an advanced driver assistance system (ADAS).
The study found that drivers preferred wider broken lane line markings, and broken lane line markings with more marking and less gap than current MnDOT practice. No specific preferences were found for contrast markings. The researchers recommend that MnDOT adopt the 12.5-foot to 37.5-foot broken lane line pavement marking pattern as it is preferred by road users and maintains the current MnDOT 50-foot cycle length. In addition, to skip line and gap dimensions, the research team also recommends MnDOT adopt 6-inch-wide pavement markings as the research results show that not only are 6-inch-wide pavement markings preferred over 4-inch-wide pavement markings by road users, they also offer a safety benefit.
