2.0 DATA SOURCES, AVAILABILITY, AND RISKS

2.1 Data Sources

The safety data test plan focuses on obtaining and analyzing crash data and incident response data on I-35W, Cedar Avenue, Highway 77, County Road 23 in Dakota County, Marquette and 2nd Avenues in downtown Minneapolis, and other key locations. Similar data will be examined for the I-394 and I-94N control corridors and region-wide trends.

The safety data test plan uses three sources of data. The first two sources are the Mn/DOT Incident Logs/Computer Aided Dispatch (CAD) system and the Minnesota Department of Public Safety (DPS) Crash Database. The third source is the Traffic.com/NAVTEQ Incident and Event Database. The data from these sources is described next, along with the advantages and limitations of each database. Data from the Mn/DOT Incident Logs/CAD system is also used in the traffic data test plan.

Mn/DOT Incident Log and/or Computer-Aided Dispatch (CAD) System. The Mn/DOT Incident Log has been used historically to record freeway incidents and incident clearance times. In August 2008, Mn/DOT transitioned to the State Patrol CAD system to collect this data. The Incident Log and the CAD system include the following data:

• incident type – crash, stall, debris, etc.;
• location – direction of travel, freeway, cross-street;
• start time and end time;
• response times for state patrol and Freeway Service Patrol;
• incident clearance time and time to clear lanes; and
• lane blocking – yes or no.

There are a number of advantages to using the Incident Log and the CAD system. First, information is collected in real-time and is available immediately. Second, data on incident clearance time and response time are available. The information from the Incident Log and the CAD system can be used to provide a preliminary snapshot of the safety-related impacts of the UPA projects and to identify potential problem locations that may occur early in a UPA project.

There are limitations to use of the Incident Log and CAD system, however. Detailed information on crashes, such as type and injury severity, is not provided. The reported crash location may not be the actual location where the crash occurred. For example, a crash occurs in a general-purpose freeway lane, but the motorist moves to the nearest ramp, which is the location recorded in the log or CAD system. Only the freeways are included in the Incident Logs and the CAD system. As a result, data on I-35W and the Crosstown Commerce Section are available, but not Cedar Avenue, Marquette and Second Avenues in downtown Minneapolis, and other locations. Historical data are available only for the operating hours of Mn/DOT's Traffic Operations, which are 5:30 a.m. to 8:30 p.m., Monday through Friday, 10:00 a.m. to 6:00 p.m. on Saturday, and 11:00 a.m. to 7:00 p.m. on Sunday.

DPS Crash Database. The DPS Crash Database includes the official crash reports from the Minnesota State Patrol and local law enforcement departments. Citizen reports are also included in the database. The list below highlights the major elements included in the DPS Crash Database:

• severity of crash – fatal, injury A, B, or C, property damage;
• crash type – struck other vehicle, struck guard rail, struck median barrier, etc.;
• crash diagram – rear-end, side-swipe, right angle, ran off road, etc.;
• location – direction, road, cross-street;
• lighting – day or night; and
• road surface – dry, rain, snow, etc.

A major advantage of the DPS Crash Database is that it includes more detailed information on the nature of the crash, including the location, than the Incident Logs or CAD System. Another advantage is that crashes are recorded 24 hours a day, 7 days a week, 365 days a year. A final advantage of the DPS Crash Database for the UPA evaluation is that Mn/DOT can request data on city and county roads, including Marquette and 2^nd Avenues in downtown Minneapolis, Cedar Avenue, and County Road 23.

The major limitation of the DPS Crash Database is the lag time in data availability. It takes approximately six months for data to become available. Given the scheduled one-year of post-deployment data collection and analysis, use of the DPS Crash Database may be limited to the first six months rather than the entire year of data.

Traffic.com/NAVTEQ Incident and Event Database. As part of SHRP2 Project L03: Analytical Procedures for Determining the Impacts of Reliability Mitigation Strategies, TTI examined the use of incident and crash data available from public agencies and private vendors. The Traffic.com/NAVTEQ database, which is owned and maintained by the private sector, was one of the commercial databases included in the analysis. As part of the SHRP2 project, TTI obtained and examined 2006 and 2007 data for Minneapolis from the Traffic.com/NAVTEQ Incident and Event Database. The data includes the incident start time, the incident end time, and the type of incident. Data from 2008 is currently being processed.

One advantage of the Traffic.com/NAVTEQ Incident and Event Database is that it includes information on freeways and other roadways in the area. As a result, before-deployment data may be available for some roadways not covered in the Mn/DOT and DPS databases. Data can be requested and analyzed on a three-to-six-month basis in the post-deployment period.

2.2 Data Availability

Table 2-1 presents the data sources and availability of pre-deployment data and post-deployment data. Pre- and post-deployment data are available from the Mn/DOT Incident Log/CAD System for I-35W and Highway 77, as well as the control corridors and region-wide trends. Pre- and post-deployment data from the DPS database are also available for I-35W, Highway 77, the control corridors, and region-wide trends. Post-deployment data will be available for Cedar Avenue, County Road 23, and Marquette and 2^nd Avenues. Pre- and post-deployment data on incident type, incident start time, and incident end time are available on selected segments of I-35W and other roadways of interest in the I-35W corridor from the Traffic.com/NAVTEQ Incident and Event Database.

2.3 Potential Risks

There are inherent limitations and risks associated with the use of traffic crash and incident databases and safety-related analyses. First, crash and incident data are recorded by personnel at the scene. As a result, the accuracy of the data depends on individuals providing accurate and complete information. Second, even when accurate and complete information is recorded, the exact cause(s) of a crash or incident may not be apparent or known. For example, a crash in the HOT lane may be the result of driver actions in the general-purpose freeway lanes or visa-versa. Third, there is a lag time with the availability of data from some of the databases. Fourth, other factors besides the UPA projects may influence reductions or increases in crashes. Examples of these factors include traffic safety campaigns, other non-UPA improvements, and reduction in VMT due to the economic slowdown. Finally, fully examining the safety improvements of projects takes longer than the one-year post-deployment period available for analyzing the Minnesota UPA projects.

All of these factors present limitations and potential risks for evaluating the safety-related impacts on the Minnesota UPA projects. The use of the multiple databases in this test plan, along with examining crash data from the control corridors and the overall tracking of changes in crashes in the metropolitan area, helps address some of the risks. Given budget and time considerations, there are no other cost effective approaches for addressing these risks. Mn/DOT and other participation agencies will be able to use the evaluation protocol to continue to monitor and analyze crash and incident data in the future, which will provide a longer-term perspective on the safety impacts of the UPA projects.