Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 15
C H A P T E R 3 Dimensions of Data Feasibility To determine the feasibility of using data from projects based ica, and the challenge of using data from another country, this on in-vehicle video data, the accuracy of two components of data set is considered unsuitable. the data is critical. The components are (1) the video data and (2) parametric data, such as GPS data, radar-detected data, Project 4: Lane Change FOT and complementary data. It is preferable to have continu- Because this is an older data set, only the most urgent lane ously recorded and high-frequency data that are capable of change events have been converted to digital video. Significant showing the details of drivers' maneuvers, environment, and vehicle status. The existence of complementary data and the effort and time would be required to digitize the data. Conse- availability of, for example, flow data, crash data, and weather quently, this data set is deemed unsuitable for the current study. data need to be investigated. An accurate time and location Project 9: Naturalistic Driving Performance stamp that enables a proper link between vehicle data and com- plementary data is necessary to the vehicle and complementary During Secondary Tasks data sources. Figure 3.1 shows the conceptual relationship This study used a small subset of data collected during the between the data sets to be analyzed and the potential travel RDCWS FOT conducted by UMTRI. Because the RDCWS time reliability improvement measures. FOT is included in this research effort and is discussed in detail Some of the studies listed in Table 2.1 are readily suitable later, this subset of the data set is not considered further. to serve the research purpose of this study. The quality of vehicle data and complementary data of each data set are dis- Project 10: Effect of In-Vehicle Video cussed individually in this section. and Performance Feedback on Teen Driving Behavior Quality of Vehicle Data Instead of continuously recording data, the video cameras The following data sets were eliminated from the candidate in this study were trigger-activated. Therefore, this data set list after examining the data size, quality, and availability. is eliminated from further consideration. Project 1: Sleeper Berth Project 12: CICAS-V Infrastructure The Sleeper Berth project instrumented two tractor trailers The CICAS-V infrastructure study collected almost 1.5 TB with cameras and DASs. Data from only 41 drivers were used of video and radar data for approaching vehicles at several in this study. Because of data acquisition limitations at the instrumented signalized intersections. Driver image data time the data were gathered, this data set is event-triggered. were not collected in this study. The videos were installed at Consequently, it is not considered in the current study. the intersections to capture vehicle movements and thus are not in-vehicle video data. Consequently, this data set is not Project 3: Quality of Behavioral considered further. and Environmental Indicators Used to Infer the Intention to Change Lanes Project 13: Pilot Study to Test Multiple Medication Usage and Driving Functioning This data set contains only 22 participants, and the data were collected in France. Because of the small sample size, potential According to the consent forms for the study, NHTSA differences in driver behavior relative to that in North Amer- is not allowed to share the video data with other parties. 15
OCR for page 15
16 Figure 3.1. Relationship between project data sets and travel time reliability improvement. Consequently, the team would not have access to the data, (approximately 40 mi). The driving data were not collected and thus it is not considered further. using a completely naturalistic method, and the length and roadways of the route are limited; therefore, this data set is Project 14: Older Driver FOT excluded from the analysis. Data collection for this study is under way. It is anticipated Project 16: Volvo Driving Behavior FOT that the data should be available in early 2010. Although the resulting data set has great potential to be used for future This study commenced in May 2008 and is expected to last studies, it cannot be used in this study, given that the data for 3 years. It will still be ongoing by the time this research are still being gathered. effort ends. Therefore, this study is excluded. Project 15: CICAS-V Pilot FOT Data Sets After Initial Screening The CICAS-V pilot study has a relatively small data set. Only After the initial filtering, the resulting valid data sets are listed 87 drivers were recruited for this study, which involved driv- in Table 3.1. As can be seen from the table, data for Project 2 ing an experimental vehicle on a predetermined short route were collected in southeastern Michigan. For Project 5, data