Summary of Projects Funded By The Federal Highway Administration for The National Surface Transportation Safety Center for Excellence From July 2006 to June 2014

Facilitated by this project, VTTI completed needed upgrades to its computational and data management resources as well as improvements to data center operations and data analysis support software used in NDSs. The following upgrades were made:

• Established a server virtualization test environment.
• Improved operational monitoring of service uptime.
• Upgraded to general purpose database servers supporting NDS instrumentation.
• Researched project management Web servers.
• Upgraded compute cluster dispatch node.
• Acquired software licenses to support data analysis.

This study presented an evaluation of driver performance while sending a text message via handheld mobile phones and an in-vehicle texting system. Participants sent and received text messages using their personal mobile phones and the vehicle's system while driving with an experimenter on a closed-road course. The test vehicle was an instrumented 2010 Mercury® Mariner equipped with an original equipment manufacturer in-vehicle system that supports text messaging and voice control of mobile devices via Bluetooth®, which was modified to allow text messaging during driving. A total of 20 participants were tested, 11 from the ages of 19 to 34 and 9 from the ages of 39 to 51. All participants were regular users of the in-vehicle system, although none had experience with the texting functions.

Results indicated that handheld text message sending and receiving resulted in higher mental demand, more frequent and longer glances away from the roadway, and degraded steering measures compared to the baseline. Using the in-vehicle system to send messages showed less performance degradation but still had more task-related interior glance time and higher mental demand than the baseline. Using the system's text-to-speech functionality for incoming messages showed no differences from the baseline. These findings suggest that using handheld phones to send and receive text messages may interfere with drivers' visual and steering behaviors; the in-vehicle system showed improvement, but performance was not at baseline levels during message sending.

The purpose of this project was to investigate data sharing as it pertains to naturalistic driving across different countries. VTTI worked with other countries to collect and share naturalistic driving data.

As part of this project, the data sharing international policies and guidelines of the following countries were reviewed: the United States, Canada, Australia, Sweden, China, Germany, Japan, the United Kingdom, France, New Zealand, and Israel.

Driving scenarios (e.g., driving relatively straight, negotiating a cloverleaf, turning at an intersection, or decelerating for a light) affect the driving-related measures collected for vehicles. During this project, automated methods were explored to review naturalistic driving data and to classify the epochs of the data according to different driving scenarios. In this way, the variance in the data created by common driving scenarios could be parsed out earlier during the data-mining process.

To explore the potential for scenario classification of this type, researchers developed operational definitions for the different scenarios, trained a data reductionist, and conducted video reduction on 60 drivers, which determined when drivers were involved in the scenarios. In the review, approximately 1,465 epochs of drivers involved in 26 different scenarios were identified. Scenarios such as left and right turns were frequently found (more than 300 cases), whereas scenarios such as interchange merges to the left or right were found in only a single trip. These cases were divided into subsets where possible and used to guide code development testing. While the code was intended to explore a proof-of-concept for scenario classification using kinematic and other variables, an unrelated effort was found to have considerable power to classify roadway scenarios. That effort matched GPS records to digital map data. Attributes in the digital maps clearly defined the roadway scenario. The GPS road-matching approach has distinct advantages in that it provides access to numerous roadway attributes available in digital maps (e.g., functional class, number of lanes, speed limit), and it leaves the kinematic variables largely not confounded for subsequent analysis. The kinematic scenario classification method that was being pursued was surpassed by this breakthrough in associating driving with the roadway segment. Based on this, the NSTSCE stakeholders decided to conclude activities on this project.

This project built on the lessons learned regarding instrumentation during the first naturalistic motorcycle study. The first study, conducted for the Motorcycle Safety Foundation, involved the instrumentation of 100 motorcycles. In the planning phase of that study, seven motorcycle models were selected for inclusion, and instrumentation was developed to fit those specific motorcycles. While the instrumentation performed well, locating willing riders of those specific models severely limited the project's approach and recruitment. To remedy this, the cycle bracket intended to support the design of bracketry and housings that would be feasible to use on a range of motorcycle models.

The most common motorcycle models were identified using records of models from State transportation department registration lists. These models were ordered from most to least common; then, bracketry solutions were developed that could accommodate the largest number of motorcycles possible. Off-the-shelf bracketry solutions were identified for many of the most common models of motorcycles; thus, generic components were engineered that complemented the off-the-shelf products. These generic components permit rapid placement of cameras and radars on different makes and models of bikes in a number of locations. The components also permit the installer to adjust angles of the equipment according to the geometry of the bike.

The solutions were developed in prototype form, tested for fit and installation feasibility, and produced in quantity. The brackets were used in the National Highway Traffic Safety Administration 160-Motorcycle study.

The goal of this project was to organize databases and materials that would facilitate use of the naturalistic data housed at VTTI by individuals more familiar with the roadway and infrastructure domains (e.g., civil engineers) than the driver and vehicle domains.

Within this project, a range of datasets were pursued that would be of particular interest to individuals working on infrastructure design-related research. Methods were developed to associate roadway attributes with naturalistic data from the 100-car study. A proof of concept was also done with 6,000 SHRP2 files. The database contained tables with data from five State transportation departments, digital mapping software, the SHRP2 Roadway Information Database, and the FHWA Office of Highway Performance Monitoring. The tables contained information describing roadway segments, such as the number of lanes, average annual daily traffic counts for heavy trucks and light vehicles, International Roughness Index Scores, vehicle miles traveled, and functional class. Tables also identified features such as transition zones, frontage roads, bridges, tunnels, etc. These attributes were coded and organized to identify epochs of time within the naturalistic datasets in which the different roadway related descriptors apply. For example, the roughness index table identified files and timestamps within files in which participants were driving on pavement identified with a particular roughness index.