The University of Virginia Center for Transportation Studies is one of three members of the Connected Vehicle/Infrastructure University Transportation Center (CVI UTC) Consortium led by the Virginia Tech Transportation Institute. The CVI UTC supports research, education and technical transfer activities.

Their website is found here:

Some of our recent CVI-UTC projects include:

Recent project reports include:

Final UTC Research Report_Merge Management

UTC Report Infrastrucure Safety Assessment


UVA CTS Presents Four CVI-UTC Research Projects at the 9th University Transportation Centers Spotlight Conference

The University of Virginia Center for Transportation Studies (UVA CTS) will present four research projects at the Connected and Automated Vehicles: The 9th University Transportation Centers Spotlight Conference at the National Academy of Sciences in Washington, DC on November 4-5. These projects are either completed or being conducted as part of the Connected Vehicle/Infrastructure University Transportation Center (CVI UTC) and presented below:

  1. Quantifying the Benefits and Costs of Virtual Dynamic Message Signs Relative to Traditional Dynamic Message Signs Using a Case Study of the I-66 Connected Vehicle Testbed. Alona Green, David Recht, Hyungjun Park, and Brian Smith
  2. Monitoring Pavement Conditions Using a Connected Vehicle-based Application. Huanghui Zeng, Brian Smith, and Hyungjun Park
  3. Identifying Safety-Critical Events using the Basic Safety Message. Robert Kluger, Brian Smith, and Hyungjun Park
  4. Virtual Dynamic Message Signs: Using Infrastructure-to-Vehicle Connectivity to Better Inform Travelers. Jiaqi Ma and Brian Smith

More details of these four presentations are provided below.

Quantifying the Benefits and Costs of Virtual Dynamic Message Signs Relative to Traditional Dynamic Message Signs Using a Case Study of the I-66 Connected Vehicle Testbed

Alona Green, David Recht, Hyungjun Park, and Brian Smith

Dynamic Message Signs (DMS) have historically provided an avenue for State DOTs to convey information to the travelling public. However, such systems have considerable maintenance costs, and potentially can be distracting to drivers. Existing literature illustrates an emerging technology of Virtual Dynamic Message Signs (VDMS). VDMS systems have the potential to reduce maintenance costs associated with traditional, hard wired signs, and also have the ability to broadcast highly-specific messages to individual vehicles.

In order to provide a basis for advancing the VDMS application, it is necessary to quantify the benefits (and costs) of the system in a monetization study that compares VDMS to traditional DMS. This is done by performing a review of existing literature for traditional DMS systems, particularly with regard to construction, operation, and maintenance. This control case is then compared to a hypothetical VDMS rollout, using operator and vehicle testing of the system as a basis for quantifying the costs and benefits, and, in turn, comparing this to DMS systems within a uniform framework.

Both traditional dynamic message signs (DMS) and proposed virtual dynamic message signs (VDMS) have hard (infrastructure) and soft (operator) costs associated with their construction, operation, and maintenance. Existing literature (in the form of manifests from existing State DOTs) document the costs of dynamic message signs. While there is a monetization component of the study, the methodology also creates a transformation mechanism to compare existing DMS infrastructure to newly-established VDMS infrastructure in both monetized (U.S. Dollars) and non-monetized (time and materials) terms. This is analyzed within the overall development of connected vehicle infrastructure, with VDMS being a single component of a larger, more-holistic, connected and autonomous vehicle evolution.

This existing data was compared to the rollout of a new VDMS system on the IH-66 testbed. Data was measured in terms of quantifying initial startup to construct the VDMS infrastructure, as well as quantification of operation of the VDMS system. The results document the benefits of an essentially simpler system: that of a communications network that disseminates specific user information to motorists without message board infrastructure. It was determined that there is value in the specificity of the messages, as compared to the more-universal DMS postings. The hard cost, especially when considered with VDMS as a single component of connected vehicles, is found to be lower for VDMS (compared to DMS). Likewise, a debugged VDMS system has lower maintenance costs, since most of the infrastructure is “soft” (i.e. software and information management) as opposed to “hard” (i.e. message boards).

This study is being performed in concert with an operator study of VDMS benefits. VDMS applications do have measurable benefits, and likely will be part of a larger migration to connected vehicle technology, which underscores the importance of future research into some of the motorist, operator, and overall benefits to the transportation network due to enhanced communication due to Virtual Dynamic Message Signs.

Monitoring Pavement Conditions Using a Connected Vehicle-based Application

Huanghui Zeng, Brian Smith, and Hyungjun Park

Transportation agencies devote significant resources towards the collection of highly detailed and accurate roadway pavement roughness data using profiler vans to support pavement maintenance decisions. Given the need for specialized equipment and sensors in collecting detailed and accurate pavement roughness data, it is very difficult to collect this data at numerous locations in a timely and cost effective manner. This study introduces a cost-effective application that utilizes existing accelerometers and GPS in Smartphones/tablets or on-board units as the set of sensors to collect and assess pavement condition continuously.

Probe data were collected on a total of 50 miles roadway from three different types of vehicles using two tablets under naturalistic driving conditions. The International Roughness Index (IRI) were also collected on the same segments using a profile van and used as the ground truth. Two challenges were addressed through this project: a) the changing vehicle speed and b) different vehicle dynamic systems.

An enhanced acceleration-based metric was developed to identify deficient pavement sections based on probe data, including vehicle speed, accelerations, and GPS coordinates. The proposed method is able to generate consistent pavement roughness measurement under changing vehicle speeds. The enhanced metric is able to correctly identify between 80 and 93 percent of all deficient pavement sections using a logistic regression model.

In addition, this study investigated the impact of vehicle dynamic systems on the proposed pavement condition metric. The sensitivity analysis and relationship analysis based on quarter-car model simulations found that the proposed metric is most sensitive to the spring stiffness of the sprung mass and least sensitive to the loading of the vehicle. Data analysis also observed an approximate linear correlation between enhanced acceleration-based metrics calculated based on data from different vehicle dynamic systems. Finally, case study was conducted to demonstrate how to calibration data collected from different vehicles.

There is a promising potential to expend this application to a connected vehicle application, given that the Basic Safety Message (BSM) includes all data elements (GPS coordinates, speed, acceleration) that are required to calculate the proposed pavement roughness index. In other words, the entire connected vehicle fleet could be turned into “probe” detecting the condition of the whole roadway pavement network in a timely and cost effective manner.

Identifying Safety-Critical Events using the Basic Safety Message

Robert Kluger, Brian Smith, and Hyungjun Park

With the emergence of connected vehicle technology, huge amounts of kinematic vehicle data will soon be projected in the form of a standardized, Basic Safety Message. With access to this type of data, a variety of improvements can be made to the way safety studies and evaluations are carried out in research, and in practice. The first step to safety studies is identifying where the events of interest are occurring, whether it be crashes, near-crashes, or even evasive maneuvers. The volume of data makes it necessary to have an automated process in place to identify events that are likely to be safety-critical, or have at-least the potential to result in a crash. In this paper, a 5-step, heuristic, algorithm was developed to identify safety-critical events using time-series speed and acceleration data from a set of 11 trips with crashes and 14 trips without any crashes or near-crashes. First events that have unexpectedly changes in speeds between observations are flagged. After the preliminary screening, logistic regression model was used to establish the probability that a safety-critical event occurred. The algorithm was able to achieve a precision of 0.78 and a recall of 0.64. The algorithm identified two false positives over about 12 hours of driving. This algorithm performed better than other published methods that were tested, and only relied on having the speed and acceleration of a vehicle involved in the crash.

Virtual Dynamic Message Signs: Using Infrastructure-to-Vehicle Connectivity to Better Inform Travelers

Jiaqi Ma and Brian Smith

Traveler Information Systems are designed and operated by transportation agencies to provide travelers with real time traffic information, enabling them to make better travel decisions. The most commonly used way to provide real time en-route traveler information to motorists is though Dynamic messages signs (DMS). Despite their effectiveness, they are costly and limited in the amount of information they can deliver. The wide availability of smart devices and the development of connected vehicles offer the possibility to provide traveler information not only through in-vehicle devices without incurring huge infrastructure costs, but also in a more flexible manner to selected individuals and locations without geographical constraints. To comprehensively develop and evaluate this concept, this research was carried out from three perspectives.

First, this research developed a smartphone-based VDMS (Virtual Dynamic Message Signs) prototype app to demonstrate the proposed concept and gain experience for future system design and deployment. Key technical performance of the app was evaluated and a user experience survey was conducted after participants use the app for several weeks. Many valuable insights are obtained: The developed VDMS app has satisfying technical performance from the perspective of impact on battery life, latency and location accuracy and further refinements, such as incorporating technologies such as DSRC, would make this concept easy to deploy. Also, the survey reveals a positive attitude among subjects toward VDMS in terms of both usefulness and satisfaction. The survey also indicates that drivers perceive that VDMS is a safer way to receive information and they feel more comfortable receiving information from VDMS compared with DMS. The results indicate great user acceptability and potentials for such systems to be deployed in the future.

Second, since the concept of VDMS uses in-vehicle devices, particularly relying on audible messages, this research aims to address the question whether this new way of information delivery conveys information as least as effectively as existing DMSs. A mixed repeated measure experiment was designed using a driver simulator to examine the impacts of driver age, information transmission mode, amount of information, and driving complexity on message comprehension, distraction and perceived difficulty. Forty two Participants were recruited and each of them was tested under different combinations of the three within-subject factors. Participant performance was measured in terms of message comprehension, distraction, and self-reported message difficulty level. Results reveal that VDMS generally performs better than DMS across different amounts of information and under different driving conditions, regardless of driver age. VDMS increases message comprehension by 16% under relatively complex conditions, reduces reaction time to unexpected stimuli (as measured with a reduced time-to-brake of 0.39 seconds), and makes the same messages easier to process and retain for drivers than DMS. Based on these results, it is recommended that transportation agencies give careful consideration to VDMS as a future strategy for delivering public traffic information in a connected vehicle environment.