VTechWorks is currently undergoing indexing. There may be some unusual behavior with item visibility on February 24, 2018, until the index is complete. Please contact email@example.com if there are continuing difficulties.
Integrating the Adaptive Lighting Database with SHRP 2 Naturalistic Driving Data
Gibbons, Ronald B. (Ronald Bruce)
Flintsch, Alejandra M.
MetadataShow full item record
This report details efforts to integrate the Adaptive Lighting Database (ALD) with the Second Strategic Highway Research Program (SHRP 2) Naturalistic Driving Study (NDS) and the Roadway Information Database. The ALD provides detailed in situ lighting performance data and associated safety, traffic, and roadway data for seven states: Washington (WA), North Carolina (NC), California (CA), Delaware (DE), Minnesota (MN), Vermont (VT), and Virginia (VA). The SHRP 2 database provides naturalistic driving data from a large-scale study carried out at six sites around the nation: Bloomington, Indiana; central Pennsylvania; Tampa Bay, Florida; Buffalo, New York; Durham, North Carolina; and Seattle, Washington. The RID, which was developed as part of the SHRP 2 program, provides detailed traffic and roadway information for the SHRP 2 sites. The integration of these datasets would make it possible for researchers to investigate relationships between different lighting characteristics, roadway configurations, and roadway safety. With this objective in mind, the research team developed an in-depth description of the NDS database structure, data elements, and database relationships; documented in detail the data entities, format, and content of the ALD; and developed and demonstrated two Geographic Information System (GIS) approaches for integrating NDS and ALD data. The two GIS approaches target different needs and requirements. The first approach involved data integration directly between the ALD and NDS time series data. By matching both lighting and time series data points onto the same roadway network, simple spatial joins or linear referencing mechanisms could be used to relate individual points from both datasets. The approach involved both existing and custom tools developed on the ArcGIS platform. Data in both databases could then be integrated through spatial and relational joins. The researchers used data for the State of Washington to demonstrate the approach and associated advantages and challenges. Time series data representing approximately 2,800 nighttime NDS trips were matched to the ALD roadway network. The second approach involved data integration of the ALD and NDS data based on the roadway segments in the RID. Within the NDS database, time series data points were matched to a digital map of the roadway network defined by links (uniquely identified by a LINKID), directly isolating time series data on the links of interest and eliminating the need for additional spatial processing. The RID roadway network was then matched with the ALD roadway network and each ALD roadway segment was assigned the LINKID of the corresponding RID roadway segment, allowing relational database joins to be used, which are many orders of magnitude faster than spatial joins. To demonstrate this approach, the research team used a draft version of the RID roadway data and lighting data for the State of North Carolina.