Archiving, Sharing, and Quantifying Reliability of Traffic Data

Vast quantities of transportation data are automatically recorded by intelligent transportations infrastructure, such as inductive loop detectors, video cameras, and side-fire radar devices. Such devices are typically deployed by traffic management centers (TMCs), and the data used for operational studies; however, such data are also highly valuable for transportation planning and other applications. This project considered how such data can best be stored and managed to accommodate multiple users, and multiple types of detector technologies. A modular system is developed, allowing data from multiple TMCs to be collected, translated into a common format, and placed in a central archive. Additionally, a novel method for quantifying data reliability is described, as error detection is critical when managing large quantities of data. Multiple techniques are also described for imputing missing data, or correcting erroneous data. Issues related to implementation are also discussed, along with innovative detector technologies that may be deployed in the near future, and thus must be considered when developing a flexible archival system.