Link-level travel time measures-based level of service thresholds by the posted speed limit

Abstract

Macroscopic parameters, such as density, volume-to-capacity ratio, and speed, are traditionally used to assess the level of service (LOS) of freeway/expressway or arterial road links. Capturing real-world data pertaining to these macroscopic parameters for all the links in the transportation network is expensive, tedious, and challenging. Additionally, density and volume-to-capacity ratio are not easily perceived by the commuters. Recent trends indicate that large-scale travel time data can be collected anonymously and without the use of field technicians, through global positioning systems (GPS), sensors, smartphones with apps, and other technological devices. The focus of this research is to develop link-level travel time measures-based LOS thresholds for urban areas, using large-scale travel time data from a private data source. The posted speed limit for each selected link in the city of Charlotte, North Carolina was integrated with the raw travel time data for the corresponding link. Travel time measures such as the average travel time, the 95th percentile travel time (planning time, PT), the planning time index (PTI), and the buffer time index (BTI) were computed and categorized by the posted speed limit of each selected link. The relationships between estimated speeds from the regional network model and computed travel time measures were then examined to develop LOS thresholds at link-level by the posted speed limit.