With increasing frequency, the public is exposed to news coverage announcing the autonomous vehicle (AV) revolution and the ways in which mobile devices, sensors and other connected vehicle (CV) technologies are redefining the automotive world. As engineers and other specialists contend with the technical aspects of AVs and CVs, society grapples with questions about safety, reliability and other potential effects on personal mobility.
Connected vehicles leverage a number of communication technologies, often sensors and mobile devices, to communicate with the driver, other vehicles and roadside infrastructure. Autonomous vehicles, often referred to as driverless or self-driving cars, are capable of sensing their surrounding environment and reacting without human input. Together, CVs and AVs represent the leading edge of innovative solutions to address congestion, environmental concerns and traffic safety. However, they also represent complex public policy issues related to social, economic and consumer impacts.
Researchers within the MATS UTC consortium are seeking to understand how ever-evolving CV and AV technical advances can be leveraged for more efficient and safe use of the built environment. CVs and AVs represent the opportunity to gather real-time mobile data, revealing important traffic information and driving behaviors not available through existing monitoring devices such as stationary cameras. Many of these research efforts are being applied to existing traffic issues, such as safety and movement through intersections, and environmental concerns, such as fuel economies and emissions.
Examples of these research efforts include:
Researchers at the University of Delaware and Morgan State University are using connected vehicle technology to optimize a vehicle’s control system in real-time to reduce congestion, improve fuel economy and reduce emissions. Using hybrid buses operating at the University of Delaware, the team is studying how intelligently integrated components can respond to both routine and atypical traffic situations, resulting in optimized traffic control and vehicle fuel economy.
Communication between a traffic signal controller and a vehicle equipped with a global positioning system (GPS) and communication hardware provides a research team from Virginia Tech and Morgan State University with sufficient information (vehicle position, vehicle speed and signal phasing and timing data) to compute fuel-efficient speeds. The team is leveraging this communication, developing Eco-Speed Control algorithms for buses using predictive energy estimation models. These models identify optimum speed profiles using information from surrounding vehicles and upcoming signalized intersections. The goal is to predict the most efficient speed to move a bus through an intersection and reduce ‘stop/go’ behavior, a key reason for inefficient fuel economy.
Collecting high-resolution speed and acceleration data is now feasible with mobile consumer devices such as smartphones. Smartphones are equipped with sensors capable of recording vehicle performance data at a very fine temporal resolution in a cost-effective way. Researchers at Old Dominion University used this mobile sensor data to identify unsafe driving patterns and quantified the relationship between these driving patterns and traffic crash incidences. The models with microscopic traffic measures were shown to be statistically better than traditional models that only control for roadway geometry and traffic exposure variables.
Research conducted at Virginia Tech seeks to develop a dynamic speed harmonization application (SPD-HARM) that makes use of the frequently collected and rapidly disseminated multi-source data drawn from connected travelers, roadside sensors, and infrastructure. Using the connected vehicle environment, the research team is developing systems and algorithms to generate traffic condition predictions, alternative scenarios and solution evaluations in real-time. The goal is to reduce crashes, whether due to speeding, poor visibility, inclement weather or construction activities.
With growing use of connected vehicles equipped with communication technologies such as GPS to communicate with the driver, other cars and roadside infrastructure, researchers at Old Dominion University, Virginia Tech and Marshall University are exploring optimization of current adaptive signal control technology to estimate queue length and develop enhanced signal coordination through communication with CV sensors. The research focuses on Traffic Responsive Plan Selection (TRPS), an underutilized adaptive control product enabling the selection of pre-programmed traffic signal timing plans based on vehicle demand observed from selected vehicle detectors along a signalized corridor.
Autonomous vehicles are typically equipped with LIDAR (light detection and ranging remote sensing technology) or other similar sensors to detect obstacles in the surrounding environment and can be a means to track other vehicles in adjacent lanes. At Old Dominion University, LIDAR is being used to estimate traffic flow parameters along the path of the autonomous car from point-cloud data. New algorithms and models are in development to extract traffic flow information from raw LIDAR data, enabling real-world data collection for safety studies and estimations of traffic flow and driving behavior.
In an automated environment, it is possible bikers and pedestrians will be safer due to improved braking technologies. However, safety may be negatively impacted if drivers, cyclists and pedestrians over-rely on automated technology. If, for example, pedestrians and cyclists assume AVs will ‘automatically’ stop for them, then there may be increases in unsafe walking and cycling behaviors such as jay-walking or failing to use designated bike lanes. Researchers at Virginia Tech and the University of Virginia are using semi-structured interviews with various stakeholders to develop planning guidelines for walking and cycling as society transitions to an automated fleet.
These research efforts are contributing to our understanding of the challenges associated with dynamic traffic conditions in an automated and connected environment. Most importantly, these projects are using real-time traffic data to develop new approaches, such as optimized traffic signals and other traffic responsive systems, to reduce fuel consumption, improve safety, and minimize congestion.