What are performance measures?

Performance measures are objective ways to determine the degree of success a project, program, or initiative has had in achieving its stated goals and objectives.  In other words, they are ways to track progress.  In this case, performance measures are what we use to track the region’s progress in reducing and managing congestion.

What are the basic considerations for selecting performance measures?

Performance measures should be easily understood by the public, staff, and elected officials so they can be incorporated into the overall transportation decision-making process.  They should be sensitive to various modes of transportation such as freight and transit.  They should also be sensitive to peaking characteristics and the amount of time that congestion is experienced during the day.  Performance measures should be sensitive to congestion mitigation strategies so discernable changes can be detected.  Also, it is useful to have performance measures which can be modeled and forecasted in order to estimate future congestion levels.  However, the most significant constraint in selecting performance measures is the availability of data and the resources available to undertake data collection.  A comprehensive series of performance measures will not be useful without reliable data to back it up.

[ top ]

What performance measures are used in Southwestern Pennsylvania?

SPC’s Congestion Management Process uses travel time, speed and delay as its fundamental Performance measures. These building blocks are then used to calculate additional performance measures such as Delay per Vehicle per Mile, Total Delay, and Total Delay per Mile.

[ top ]

How is travel time and speed data collected?

Travel times and speed data are collected by the “floating vehicle” method, in which a vehicle moves with the flow of traffic during morning and evening peak hours.  An on-board computer captures travel times, average speeds, and distance measurements, which are then downloaded for analysis.

[ top ]

How is delay calculated?

Before travel time runs are performed, staff records the posted speed limit within each segment of the corridor so an ideal travel time (the amount of time it would take to traverse a given roadway segment at the posted speed limit if there were no interference) can be calculated.  The measured travel time data collected in the field is then compared to the ideal travel time to determine delay.

[ top ]

How is average speed calculated?

Average speed for each roadway segment is provided by the computer software used during data collection.  Once this data is downloaded, an average speed is calculated for the corridor.

[ top ]

How are other performance measures calculated?

Since segments and corridors vary in length, it is useful to factor distance into congestion calculations.  For example, a motorist who encounters 5 minutes of delay in a 5-mile corridor does not experience as much congestion as a motorist who encounters 5 minutes of delay in a 1-mile corridor. Therefore,  Delay per Vehicle per Mile is another useful performance measure. Delay per Vehicle is the amount of time it takes to traverse a given roadway segment minus the amount of time it would take to traverse that roadway segment at the posted speed limit if there were no interference. Delay per Vehicle per Mile is Delay per Vehicle divided by the length of the roadway segment.

Because different roadways carry different amounts of traffic, it is also useful to consider Total Delay as a performance measure.  Total Delay considers the number of vehicles that are impacted by delay.  For example, a roadway carrying 500 vehicles in the peak hour with five minutes of Delay per Vehicle per Mile is generally not as critical a congestion problem as a roadway with 5 minutes of Delay per Vehicle per Mile that carries 5,000 vehicles in the peak hour. Total Delay is calculated as Delay per Vehicle multiplied by the peak hour volume in that roadway segment. It is important to note that SPC’s CMP considers AM and PM peak periods, so Total Delay refers only to peak times and not to the entire day.

As indicated previously, since segments and corridors vary in length, it is useful to have additional performance measures that factor in distance.  Therefore, SPC’s CMP also considers Total Delay per Mile, which is calculated as Total Delay divided by the length of the roadway segment.

Each of the above mentioned performance measures are calculated for both AM and PM peak periods.  Two additional measures are utilized to look at the overall congestion levels in each corridor: Total Peak Hour Delay and Total Peak Hour Delay per Mile.  These measures are a total of the congestion in the AM and PM peaks.

For more information on data collection and proper data usage, please see the Technical Notes page.

[ top ]

Where can I find the detailed corridor data collected through SPC’s Congestion Management Process?

Check out the Performance Monitoring & Data section of this website.

[ top ]

How do I interpret the information provided for individual corridors?

The following graphs are samples illustrating the type of information that is available for each corridor.  These graphs illustrate Delay per Vehicle per Mile during the morning peak hour.  The first graph is for southbound traffic; the second graph is for northbound traffic.

As shown in these graphs, during the AM peak the Delay per Vehicle per Mile in both the northbound and southbound directions is greatest between Saxonburg Boulevard and Kittanning Street.  The fact that Delay per Vehicle per Mile, which is a measure that does not explicitly factor in traffic volumes, is significantly greater between Saxonburg Boulevard and Kittanning Street than in other segments indicates that targeted operational improvements may be an appropriate congestion mitigation strategy for this segment.

Note: Graphs with blue bars indicate Northbound or Eastbound travel and graphs with cranberry colored bars indicate Southbound or Westbound travel.

[ top ]

What about performance measures for non-recurring congestion?

Gathering data and evaluating performance measures for non-recurring congestion is significantly more difficult because of its inherent unpredictability and the need for virtually continuous data.  One source of non-recurring congestion for which consistent data is available is reportable crashes.  A crash analysis, including crash rate and severity index, is provided for each CMP corridor in the Performance Monitoring & Data section of this website [ disclaimer ].

As technology changes and additional sources of data are readily available, from cell phone probe data and automatic vehicle location (AVL) data for example, SPC will continue to investigate the potential for integrating additional non-recurring congestion performance measures into the CMP.

[ top ]

For more information about our Congestion Management Process, contact Doug Smith at (412) 391-5590 x327 or e-mail dsmith@spcregion.org.

[ top ]