PennDOT completed a study of INRIX and Waze data to assess the value of monitoring real-time crowdsourced data to improve traffic incident management operations, as well as the initial improvements PennDOT has made based in the findings.
PennDOT analyzed data for a year on its core roadway network, including over 15,000 crashes, INRIX congestion events, Waze Connected Citizen Incidents, and PennDOT Road Condition and Reporting System (RCRS).
The analysis showed crowdsourced data is essential to improve incident detection. INRIX detected 81.2% of the incidents, more than the other methods. INRIX detected high congestion crashes first at 49%, followed by Waze with 27% and traditional methods at 24%.
Using INRIX data, the analysis also generated information regarding the scale and causes of congestion for secondary crashes and work zones. For secondary crashes, 75% of the time queues were present for 15 minutes or more, and 40% for an hour or more. Twenty-two percent of the secondary crashes happened 5 miles or more upstream of the primary crash. In work zones, 24% of the crashes were 2 miles or more from the start of the work zone and 85% of the work zone-related crashes had permanent DMS within 5 miles upstream the work zone.
Leveraging the results of the analysis PennDOT made several modifications to their Regional TMC operating procedures. To reduce secondary crashes, operators enter queue protection messages on DMS and contact service patrols before entering into the incident management system. Also, messages on DMS are deactivated only after the queue is cleared instead of when the incident is cleared. PennDOT is also working toward automated DMS messaging to improve driver alerting.
Other state DOTs can replicate the PennDOT methods to improve their statewide traffic incident management. For the full case study, Pennsylvania DOT Using Crowd Sourced Data to Assess and Improve Statewide Traffic Incident Management, click here.