Operations Performance Measures Paul Szatkowski June 5, 2014
Operations Performance Measures Agenda Program History & Development Results & Recent Actions On-going Activities Good morning I would like to thank ITS Virginia for providing us an opportunity today to discuss the Operations Performance Measure program. My presentation today will not be a theoretical or academic talk. Instead, I will be sharing our lessons learned about establishing our performance measurement program, the results of the program and where we plan to take it.
Operations Performance Measures Program History & Development VDOT Dashboard Measures Congestion – Level of Service Travel Times – Commuter Routes (NoVA HOV) Hours of Delay – TTI Urban Mobility Report Incident Duration Performance Measurement is not new to VDOT. While there has been a recent surge of activity and interest in operations performance measures, it is important to recognize that VDOT has been one of the leaders in sharing performance metrics with the public through our Dashboard program. Dashboard was originally launched in 2003. Today, we are using Version 3.0. We are currently reporting various performance measures such as congestion, travel time, hours of delay and incident duration. Our current measures are good. However, as programs mature, lessons learned emerge, additional data (sources) and tools become available and analysis capabilities improve. For this reason, we took a critical look at these measures.
Operations Performance Measures Program History & Development One concern about using Dashboard is shown in the hours of delay statistic. Currently we are using a statistic prepared by the Texas Transportation Institute’s Urban Mobility Report. This statistic is very comprehensive. It includes multiple interstate and primary roadways within three urban regions; DC, Virginia Beach and Richmond. However, the statistic is difficult for VDOT to use as a benchmark. First, the report is prepared every 12 – 18 months. It is difficult for VDOT to determine our interim progress. Second, the statistic includes roadways that are not managed by VDOT. For example, the Washington DC statistic includes roadways within the District of Columbia and Maryland. In Richmond, we contacted TTI and learned that about ¾ of the congestion in greater Richmond is located on arterial highways. Also in Richmond, Henrico County is one of two counties that manage their own road network. Together, these examples show that there was a need to develop a new set of operation performance measures for the organization.
Operations Performance Measures Program History & Development VDOT – Developed Interstate PM We are here FHWA Rule Making Ends FHWA Implementation – NHS PM While this slide is out of date, it shows that when MAP-21 was introduced, there was a plan for states to provide comments about the proposed operations performance measures in 2014. We decided a year ago that it would be in our best interest to further develop our operations performance measures so we could provide feedback to the US DOT and to be prepared for this change.
Operations Performance Measures Program History & Development Travel Time Reliability (80th percentile) Vehicle Hours of Delay Last year, when VDOT started evaluating our operations performance measures, we noted that other agencies were assessing the upcoming MAP-21 program. We noted that there was consistency in the recommendation to consider using two primary performance measures; travel time reliability at the 80th percentile and vehicle hours of delay. These two items will become our primary operations performance measures. Together, they address both the frequency of congestion (through travel time reliability) and the magnitude of events (through vehicle hours of delay). Urban – VHD Non-Urban - TTR
Operations Performance Measures Program History & Development While vehicle hours of delay and travel time reliability report the current situation of congestion, they are only lagging indicators. Alone, they do not tell what contributed to any increases or decreases in congestion. For this reason, we decided to prepare additional measures to use as tools for our program. We felt that these measures should be related to causes of congestion that the operations program could impact. The slide you are now looking at shows some contributors of unreliable travel or congestion. While the statistics will vary from report to report, the pie chart shows that besides capacity issues, traffic incidents, work zones, signal timing and special events contribute to half of the causes of congestion. The impact in these areas can be mitigated through operations programs such as deploying quick clearance programs, using auto-detection technologies, providing travel time information, etc. For this reason, we looked into creating set of secondary operations performance measures.
Operations Performance Measures Program History & Development Available Data – Established Measures Incident Duration HOV Lane Performance Future Efforts Work Zones Weather While we would like to have measures for all of the causes of congestion, the availability of existing data, or the consistency of the data are limiting factors. When we started the performance measurement program, we had good data to report on incident duration and HOV lane performance statewide. These two items became our first secondary measures. Our partners in Traffic Engineering and the Virginia Center for Transportation Innovation & Research were conducting a report for the development of work zone mobility performance measures. Additional work is needed in this area before we can develop measures.
Operations Performance Measures Results & Recent Actions Introduced / Revised Operations Performance Measures Vehicle Hours of Delay Incident Duration HOV Lane Performance Incorporated Performance Measures in Operations Planning Programs Smart Roadway Technology At this point, I would like to share with you the results of our performance measurement program and how this program has affected operations planning activities, particularly the smart roadway technology program.
Operations Performance Measures Results & Recent Actions Target FY14 (July-Oct.) FY13 Travel Time Reliability Index Vehicle Hours of Delay TBD 8,951,291 8,089,879 Incident Duration* 125* 129* HOV Performance 90% 12 of 14 8 of 10 This slide shows the FY14 first quarter results. Throughout the year, our division collects and analyzes data regarding the VHD, incident duration, and HOV performance for interstate highways only and compares the results to the same time period from the previous year. Our team has developed tools for the regions to conduct drill downs on the data to help identify the causes of any spikes in congestion or incident duration. As shown on this slide, the top series of bar charts compare the VHD in each District from FY 14 to FY 13. As expected, our Northern and Hampton Roads regions experience the most VHD, followed by our Richmond and Fredericksburg Districts. The second and third series of bar chart show the number of incidents with the average duration length of incidents which take over 30 minutes to clear. These charts are showing that the while the number of reported incidents are increasing, likely due to expansion of SSP which are reporting the events, the average duration length is dropping consistently throughout the state. The final row demonstrates the performance of HOV lanes which can offer more insight on the performance of the corridor. Our findings show that HOV lanes are consistently operating at speeds of 45 mph or greater for all segments except on I-66 outside of the beltway. These reports are prepared every quarter and discussed by our management teams to determine what corrective actions or strategies should be developed to improve our performance. Vehicle Hours of Delay data source = RITIS Vehicle Probe Project - All Interstates (except Lynchburg = US29, US58, US360, US460 ) Incident data source = VaTraffic Incidents and Short Term Weather on Interstates with a duration > 30 minutes (excluding Amber Alerts and Congestion) HOV Performance = HOV operating at speeds of 45mph or greater 90% of the time .
Operations Performance Measures Results & Recent Actions – Vehicle Hours of Delay Established delay thresholds Established interstate segments Introduced results Developing targets To go into more detail about each measure, introducing the Vehicle Hours of Delay measure was the most challenging. We need to reach a lot of consensus with our management team to agree on what point to we declare that there is a delay and how to report this information. We first went through an education process about this measure. We agreed to focus on interstate highways by District. We also determined that we considered delay to begin when traffic moved 20 mph below the free-flow speed, not posted speed. We also had to develop internal tools to analyze the results made available through vehicle probe data. The slide you are looking at shows the results of a recent analysis comparing VHD from 2011 – 2013. We have found the VHD increased 5% between 2012 & 2013. The limited data for 2014 is relatively similar for 2013. (Note the quarter by quarter analysis on the right). This analysis is being used as VDOT considers an performance target for this measure.
Operations Performance Measures Results & Recent Actions – Incident Duration Revised criteria (FY15) > 30 minutes Travel lanes cleared Median duration Disabled vehicles, vehicle crashes, vehicle fires and tractor trailer crashes Introduced results Developing targets Over the past year, after reviewing the results of the incident duration measure, VDOT made some changes to this measure. The most significant change is that we are now looking at the travel lanes cleared versus the entire event duration. We are continuing to collect and monitor that statistic, but our primary reporting will be focused on travel lanes cleared. We have also decided to focus on median duration versus average duration to mitigate the influence of one or two significant events. Finally, we aligned our program to match the Dashboard requirements of reporting only events taking over 30 minutes. Again, we have data and can monitor other events, but our focus is on events that we can impact, such as 30 minute events. We have recently analyzed our data and as shown in the bar chart, there has been a steady reduction in the median duration of incidents from 2010 to date. We are using this information to set a performance goal which could be a continued five minute reduction in five years.
Operations Performance Measures Results & Recent Actions – 2014 Program Changes Adjusted safety service patrol routes Promote quick clearance Complete TIM training Review and update allowable work hours As I mentioned earlier, every two to three months, VDOT management reviews the results of our reported performance measures and discusses strategies to improve them. This slide summarizes some of the feedback from recent meetings. Some recent strategies used to improve the operations performance measures have been focused on incident management and work zone management. Safety service patrol routes have been adjusted with respect to coverage areas and hours. Training programs such as Traffic Incident Management have been expanded. Emphasis has been placed on promoting quick clearance. While these programs existed before the performance measure program was expanded, the performance measure program continues to place emphasis on them and facilitate decision making in these program areas.
Operations Performance Measures Results & Recent Actions – 2014 Program Changes One interesting outcome that has emerged is shown on this slide. Since we are reporting VHD and incident duration, our team has developed tools to merge data together to conduct analyses showing what event contributed to congestion. The slide here shows one example. In October 2013, there was a spike in the vehicle hours of delay on I-66 in our Culpeper District. A drill down was able locate the specific date and location of a work zone event on October 4th. This work zone was necessary, but it did emphasize the impact and may offer lessons learned for future events. Our office has recently made this resource available to our regions.
Operations Performance Measures Results & Recent Actions – Planning Performance Measures is a basis for the Smart Roadway Technology master plan. Smart Roadway Technology projects were developed to improve travel time reliability and reduce congestion via Technology gap closures Advanced technologies ITS equipment life cycle replacement Tunnel traffic management Arterial operations The impact of the operations performance measurement program was immediately felt on the operations planning program. We know that MAP-21 will require agencies to monitor their progress. We also know that there have been increased expectations for us to justification the allocation of resources and to show the benefit of our projects. The performance measures are just one set of tools we can use. One application of our performance measures occurred earlier this year when we developed a master plan for the ~$100M smart roadway technology program. As you may remember from our briefing at the ITS Legislative Reception, the Smart Roadway Technology program includes funding for projects or programs that reduce congestion, improve mobility, improve safety, provide up-to-date travel data, or improve emergency response. The funding can improve key corridors, but not the entire state. For this reason, we chose to develop projects and allocate funding on our most congested corridors. For example, I-95 is getting the majority of the SRT funding (40%). Other corridors such as I-85, are not being considered at this time.
Operations Performance Measures On-going Activities Introduce Travel Time Reliability (80th percentile) performance measures Expand performance measures from interstates only to all limited access highways Establish continuous data accuracy reviews Develop Work Zone data collection and analyses Develop predictive planning tools In the immediate future, our team will continue to advance the operations performance measure effort. Travel Time: Introduce concepts, then results, then targets. We have completed a study to determine how we will segment the interstate network to calculate travel time reliability. Limited Access: Evaluate data sources for non-interstate routes. Work Zones: Work is needed to get access to the existing, detailed data sources as well as to standardize the operating and reporting procedures. Predictive Planning Tools:
Operations Performance Measures Paul Szatkowski VDOT – Operations Division Paul.Szatkowski@VDOT.Virginia.gov (804) 786-8755