Project Name

Infrastructure Pavement Assessment and Management Applications Enabled by the Connected Vehicles Environment Research Program – Phase I: Proof-of-Concept

Research Team

Brian L. Smith, PE

Huanghui Zeng

Hyungjun Park


Connected Vehicles/Infrastructure University Transportation Center: Connected Vehicles/Infrastructure University Transportation Center

Project Dates

September 2012 - 2015

Project Description

The objective of this collaborative research program is to develop prototypes and conduct a field test of system level applications of a connected vehicle pavement condition measurement system. This will allow the research team to investigate (1) different approaches to a connected vehicle pavement measurement system; and (2) determine the optimum procedures for collecting, processing, aggregating, and storing the data to support engineering and management decisions. The UTC Virginia test bed will be used to support this work, and it will build upon previous work funded by the Cooperative Transportation Systems Pooled Fund Study (CTS PFS) and the Mid-Atlantic University Transportation Center (MAUTC). Specific objectives of this program can be summarized as follows: 
1. To gain experience in a system-level pavement condition measurement applications to
determine feasibility.
2. To compare a DSRC versus a smart phone based approach to this application.
3. To investigate the utility of the data produced for supporting pavement/asset management decisions (connected vehicle experts at UVA CTS will work collaboratively with pavement experts at VTTI).

Photos and Media


The research program has been developed as a two-phase effort. This proposal covers Phase I, which will include the refinement and deployment at the Smart Road of the roughness approach, and a proof-of-concept validation of the friction measurement application. The second phase (not included in this proposal) is planned to include the deployment of the ride quality applications at the more complex NOVA CV Testbed, the validation of the friction measurement application and a comprehensive evaluation of the novel pavement assessment approach.
Task 1 Refine measurement concepts (VTTI)
In this task the team will investigate the best approaches to derive roughness and friction measures from vehicular sensors. This will involve looking at previous work and modifying/enhancing the algorithms as necessary for fundamental measurement approach.
Subtask 1.a. Roughness
Roughness (or ride quality) is currently measured using the International Roughness Index (IRI). The IRI is the accumulation of the suspension travel for a simulated quarter-car vehicle model as it moves across a given length of road, usually expressed in inches/mile (mm/km) obtained using a simple quarter car model as shown in Figure 1.
One of the underling hypothesis of this project is that similar (and even more representative) information can be obtained from regular vehicles travel on the road. Preliminary work suggests that this is possible; Figure 2 compares the smoothness profile and vehicle body acceleration measurements performed at the Smart Road. Similarities between the two profiles can be observed (Figure 2). The Vertical green lines delineate pavement sections with different roughness while red circles identify high localized peaks. Both the roadway profile and vehicle vertical acceleration profile identify the same sections and peaks. Pearson’s correlation between the two signals was calculated as 0.50. This value, although not high, warrants further investigation of the relationship between the two signals.
Subtask 1.b. Skid
Another key pavement performance indicator is the level friction provided by the pavement because it is critical in cases where vehicles have to respond to challenging circumstances. In this subtask, we will evaluate different methodologies to extract friction information from connected vehicles. Tire-Road friction Estimate (TRFE) is an intensive research area since friction information is needed during the design of many vehicle control systems (4). There are numerous TRFE approaches that have been investigated and even evaluated experimentally. The various technical approaches proposed in the literature will be carefully catalogued, reviewed, and evaluated during this task. The most promising approach based on their practicality, robustness, and functionality will be selected for development and testing.
Task 2. Test measurement components on the Smart Road UTC CV Testbed (VTTI)
In this task, the research team will test the concept at the Smart Road CV Testbed to validate it based on experimental data. The Smart Road facility will be evaluated repeatedly using one of the instrumented vehicles (i.e., a research vehicle instrumented with the most recent generation data acquisition system [DAS]). The data gathered will then be compared with the information obtained from the extensive pavement surface condition collected through the Virginia Surface Properties Consortium at the Smart Road. The data accumulated by the vehicle and the type of information that can be obtained from that data are summarized in Table 1. This controlled experiment study will allow us to match the pavement condition and kinematic signatures exactly. Preliminary results suggest that under these controlled conditions, an ideal level of agreement between the pavement profile and probe vehicle dynamics can be obtained. 
Task 3. Refine system level design concepts (UVA)
In previous work conducted under the MAUTC program, UVA CTS explored alternative system level design concepts for pavement assessment data collection to support pavement management. Two fundamental concepts were identified – one using DSRC infrastructure and one based on commercial wireless networks (referred to as the smart phone approach). These are illustrated in Figure 4. 
In this task, the team will consider these approaches within the framework of the Connected Vehicles UTC. The team will use existing naturalistic driving data to use a data-driven analysis technique to determine if alternatives to the concepts are required. Finally, the team will also extend the concepts to support real-time collection of pavement friction data for use in weather operations. The time critical nature of this application will necessarily require a different approach – and may impact communications technology used. At the completion of thistask, concepts will be defined to support both real-time and non real-time applications.
Task 4. Prototype system applications (UVA)
Based on the resulting concepts of Task 3, it is likely that the team will be able to explore the effectiveness of the concepts using data collected through the VTTI instrumentation package. In other words, it will not be absolutely necessary to prototype all of the concepts in order to conduct the feasibility assessment of this project. In this task, based on a careful analysis of data collected through instrumentation and requirements of the applications, the team will chose a single concept to prototype in the UTC testbed in order to demonstrate and test the pavement applications. At this time, it is likely that the team will use Concept 2 described in Task 3 above. This will allow for the most data to be collected over the widest range of roads and conditions.
The prototyping will involve the following steps 
Document system requirements
Complete system design 
Creation of system software

Task 5. Final Report (VTTI/UVA)

The team will compile the results of the effort in order to create a comprehensive final report. In addition, the team will actively publish findings in transportation conferences and journals in order to encourage wide dissemination of the research.

Comments are closed.