Data-Driven Identification of Candidates for Operational Improvement

Project Details









Iowa Department of Transportation

Principal Investigator
Christopher Day

Affiliate Researcher, CTRE

About the research

The Safety Improvement Candidate List (SICL) process has been used by Iowa Department of Transportation (DOT) for nearly 20 years as a method of identifying locations where investments should be made. These determinations are made from analysis of crash data that consider both severity and frequency. A similar approach would be desirable for identifying locations for improvements to traffic operations: an Operational Improvement Candidate List (OICL). A challenge in compiling such a list is that operational data collection is more difficult to obtain than crash data. The emergence of new data sets offers opportunities to develop new approaches to analyzing traffic operations on a widespread basis.

Recent research on the effectiveness of adaptive signal control has suggested the use of segment speed data as a screening tool. This type of data can be very effective at evaluating corridor operations. Intersections can be more challenging to assess, since not all movements at intersections will be captured with segment speed data. However, corridor travel times and travel time reliability can be characterized with this type of data, taking into account variation by time of day, and corridors can be ranked using a variety of metrics. Furthermore, this type of data is already being collected by Iowa DOT for assessing interstate route performance, making it an excellent opportunity to better leverage existing resources to accomplish high-level screening. These analyses will be able to identify bottlenecks in the roadway system and locations where travel time is unreliable.

This research will investigate potential uses of the identified data sets to develop a methodology for populating an OICL. The research would approach the problem on two levels. The first would be to use segment speed data to perform ranking of facilities at a corridor level. The second approach would be to examine the use of high-resolution data at signalized intersections, using locations with existing capability, and using the results of the corridor level analysis to identify sites where adding this capability may be considered or temporarily deployed.