The Use of Data in Planning for Operations: State-Of-The-Practice Review

Report Image 508 Captions

Figure 1. Chart. Data collection methods to support operations performance measures. This chart depicts the contrast of data collection methods to support operations performance measures. The first column contains three methods: fixed sensor, floating car, and vehicle probe. The second column contains sub-methods for each of the three methods. For fixed sensor, the sub-methods include single loops, oval loops, cross-fire radar, and video cameras. For floating car, there is only one sub-method that is Global Positioning System (GPS) instrumented. Vehicle probe sub-methods include toll-tag transponder, fleet GPS data, and cell phone probes. The third column includes base measurements for each sub-method. Within the single loops sub-method, base measurements include volume and occupancy. For oval loops, base measurements include volume, occupancy, and speed. For both cross-fire radar and video cameras, base measurements include volume, occupancy, and possibly speed. Within the GPS instrumented sub-method, the only base measurement is travel time. For toll-tag transponder, fleet GPS data, and cell phone probes, the only base measurement is travel time. Column 4 indicates the typical sampling parameters for each of the eight sub-methods. Columns 5 and 6 indicate freeway use and arterial use, respectively, for each sub-method. Columns 7 through 13 contain, respectively, the performance measure supported by each sub-method, including speed, travel time, TT-reliability, recurring delay, non-recurring delay, extent of congestion, and throughput. Column 14 contains the costs for each method and sub-method if they differ. The costs for fixed sensor are $7,500 to $20,000 per site, depending on availability of existing structures. The costs for floating car methods are $300 to $500 per mile. The costs for toll-tag transponder are $15,000 per site per direction (exclusive of structures). The cost for fleet GPS data is $50 to $1,000 per mile per year. The costs for cell phone probes are $500 to $1,000 per mile per year. The final column, 15, indicates the primary deployment issues for each method and sub-method if they differ. The primary deployment issues with fixed sensor methods are costs, sensor density, maintenance, and quality control. The primary deployment issue with the floating car method is minimum sampling parameters. The primary deployment issue with toll-tag transponders is the density of toll-tags and costs of equipment. The primary deployment issues with fleet GPS data are data latency and sampling density. The primary deployment issues with cell phone probes are accuracy, privacy, and business model sustainability. Return

Figure 2. Chart. Capability of analysis tools/methodologies to address M&O strategies. This chart highlights the capability of analysis tools/methodologies to address management and operations (M&O) strategies. The first column lists the tool/methodology, including BCA.net, Cal-B/C, COMMUTER Model, EMFITS, FITSEval, HERS-ST (Preprocessor), IDAS, IMPACTS, Multiresolution/Multiscenarie Methods, NET_BC, SCRITS, STEAM, TOPS-BC, and TRIMMS. Columns 2 through 9 include the following strategies: travel demand management, public transit systems, arterial traffic management, CVO, HOT lanes, freeway management systems, incident management systems, regional multimodal traveler info, and workzone management. For each of the 14 rows indicating the tool/methodology, either an opaque dot, a clear dot (with outline), or nothing is depicted where the row corresponds with each strategy column. The opaque dot indicates that the tool/methodology addresses most elements of the strategy, the clear dot indicates that the tool/methodology addresses some elements of the strategy, and a blank cell (without either dot) indicates that no elements of the strategy are addressed by the tool/methodology. Return

Figure 3. Chart. Data needs for each model type to analyze TIM strategies. This chart depicts data needs for four types of models to analyze traffic incident management (TIM) strategies. The first column indicates the model type, including sketch planning models, deterministic (HCM type macroscopic) models, mesoscopic simulation models, and microscopic simulation models. These correspond with four additional columns, labeled as network, demands, TIM, and calibration. Where each of the four model types intersects with each column there is a description pertaining to how they relate. For the sketch planning models, the network is “minimal, generally deals with regional (VMT) and (VHT),” the demands are “minimal, generally regional VMT,” TIM is “general categories of TIM strategies with no specific implementation details,” and calibration is “not applicable.” For the deterministic (HCM type macroscopic) models, the network is “link and intersection-specific lane geometry, speed limits, controls (signal timing),” the demands are “link and intersection-specific hourly demands by vehicles type (usually just peak hour),” TIM is “incident type (number of lanes blocked, specific links affected, and average duration). Strategies to be tested and expected effects on average lane blockage durations,” and calibration is “not generally done.” For mesoscopic simulation models, the network is “same as the deterministic HCM,” the demands are “OD tables by hour of day for peak periods,” TIM is “same as for deterministic HCM,” and the calibration is “observed flows and link speeds.” The microscopic simulation model network is “same as HCM plus signal detector locations, signal controller settings, and turn pocket lengths,” the demands are “same as mesoscopic or link and intersection-specific demands by vehicle type,” TIM is “incident start/end times, longitudinal location within link. Expected effect of strategies on specific incident duration,” and calibration is “observed flows and link speeds.” Return

Figure 4. Chart. Sample data requirements for AMS for ICM. This chart lists the sample data requirements for analysis, modeling, and simulation (AMS) and integrated corridor management (ICM). There are five columns labeled network, travel demand, traffic control, transit, and ITS elements. Sample data requirements for network include link distances, free-flow speeds, geometric-freeways (with additional sub-requirements), geometric-arterials (with additional sub-requirements), and parking (with additional sub-requirements). Requirements within travel demand include link volume, traffic composition, on-and off-ramp volumes, turning movement counts, vehicle trip tables, person trip tables, and transit ridership. Traffic control sample data requirements include freeways (including ramp metering as well as and mainline control with additional sub-requirements for both), and arterials (including signal system description, transit signal priority system, and emergency preemption system, each with additional sub-requirements). Return

Figure 5. Flowchart. The eight-step CMP framework used by RTC. This flow chart depicts the eight-step congestion management process (CMP) framework used by the Regional Transportation Commission of Southern Nevada. The eight steps are: (1) Develop Congestion Management Objectives; (2) Identify Areas of Application (3) Define Systems/Network of Interest; (4) Develop Performance Measures; (5) Institute System Performance Monitoring Plan; (6) Identify/Evaluate Strategies; (7) Develop Congestion Management Objectives; and (8) Monitor Strategy Effectiveness. Return

Figure 6. Screen capture. Nevada FAST Web site’s interactive dashboard. This screen capture of the Nevada Freeway and Arterial System of Transportation Web site depicts an interactive dashboard. It contains a map on the left side and four boxes on the right. Each box holds a graphical representation of a performance measure. The map on the left side displays roadways around Las Vegas, NV, with two interstates color coded to represent the speed traffic is traveling. The map indicates that traffic is generally moving at freeflow conditions. The four boxes on the right each graphically display either speed or congestion. Each box resembles a dialog box with tabs; the user can click between the tabs. The upper left box contains a graph of daily peak speeds. Tabs indicate the user may select a.m. or p.m. periods. The upper right box contains time of day speeds and also allows the user to select a.m. peak or p.m. peak. The lower left hand box displays freeway average speed. Both a.m. peak speed and p.m. peak speed are represented on a speedometer. The user may select from among tabs depicting past 30 days, 30–60 days, and last year. The lower right hand box indicates congestion, which is represented with a pie chart. The pie slices are color coded to represent a range from no congestion through heavy congestion. The user can select from two tabs: a.m. peak and p.m. peak. Return

Figure 7. Screen capture. Houston TranStar color-coded traffic map. This screen capture displays a map of the Houston, TX, metropolitan area with major roads color coded from yellow to green representing congestion levels. Most road segments are green, indicating that traffic is generally traveling at freeflow speeds throughout the region, although some segments indicate yellow or orange conditions, which represent areas where there is low to moderate congestion. An interactive legend allows the user to turn on and off specific layers on the map, including the locations for incidents, road closures, cameras, motion, park-and-rides, transit centers, construction, weather radar, and message signs. Both zoom function options and a save settings option are also provided. Return

Figure 8. Screen capture. PeMS online system. This screen capture displays the California Department of Transportation’s Performance Monitoring System (PeMS) page. On the right, the system depicts a map of major freeways in the San Francisco, CA, region. Each freeway is color-coded from green to yellow, orange, and red to represent freeflow conditions through slightly congested, moderately congested, and very congested conditions, respectively. The map can be adjusted by the user to depict layers that display freeway traffic conditions, incidents, changeable message signs, and arterial traffic conditions. Other tabs allow the user to view performance, inventory, and a search function. Return

Figure 9. Screen capture. MTC 511 system. This screen capture displays the Bay Area Metropolitan Transportation Council (MTC) 511 San Francisco Bay Web page. On the left, the page depicts latest news and construction alerts and 511 driving times based on the user’s trip origin and destination inputs. To the right is a map depicting directions for the user’s route as well as the conditions on regional freeways, which are color coded to reflect freeflow (green) through congested (red) conditions. The user has multiple options for moving around and using the site, including 511 tools, access to a mobile app, and a link to a commuter benefits program for employers.Return

Figure 10. Diagram. Project performance bubble chart by project type. This diagram provides project performance assessments results by project type using differently sized bubbles. Project types include road projects, transit projects, and regional programs. Bubble size represents the total annual benefits for all projects of that type. Each circle is positioned on a four-quadrant x-y axis, where the y-axis represents benefits-cost ratio (with positive ratios or benefits above the junction of the x-axis and negative ratios below it). The x-axis to the left of the y-axis junction indicates the range for which the project type has adverse impacts on plan Bay Area targets up to a rating of -10, whereas the x-axis to the right of the y-axis junction indicates the range for which the project type is supportive of the plan Bay Area targets up to a rating of +10. Highway expansion road projects have an adverse impact on targets (about -3) but provide a positive benefit-cost ratio (at about 4). Express lane networks are road projects that have an adverse impact on targets (-1) but provide a positive benefit-cost ratio (at about 5). Road efficiency projects support the plan Bay Area targets at a rating of about 3 and provide positive benefits at about 8. Freeway performance initiatives, with a benefit of about 15, and congestion pricing, with a benefit of about 49, support the plan targets at ratings of about 4 and 5, respectively. Among transit projects, transit frequency improvements for the north Bay area have a benefit ratio of slightly less than 1 but support the plan targets at a rating of about 4. Rail expansion has a benefit ratio in the 1 to 2 range but a rating of 5 in support of the plan targets. Bus rapid transit and Infill transit stations have a slightly greater benefit, with a ratio of 2, but are slightly less supportive of the targets, falling slightly behind the level of rail expansion projects. Transit frequency improvements for the central Bay Area offer the greatest total annual benefits, with a benefit ratio of about 2 and a target support rating of 7. Finally, regional programs also all support the targets and have benefits in the 0 to plus 3 range. These programs include the climate program, with a ratio of 1 and a rating of 3; maintenance, with the greatest annual benefits of all regional programs, featuring a ratio of 2 and a rating of 5; lifeline and new freedom, with a ratio of 0 and a rating of 5; bike network, with a ratio of 2 and a rating of 7; and transportation for livable communities, with a ratio of 3 and a rating of 7.Return

Figure 11. Diagram. Project performance bubble chart: all road projects. This diagram provides project performance assessments results for all road projects using different sized bubbles. Bubble size represents the total annual benefits for all projects of that type. Each bubble is positioned on a four-quadrant x-y axis, where the y-axis represents the benefits-cost ratio (with positive ratios or benefits above the junction of the x-axis and negative ratios below it). The x-axis to the left of the y-axis junction indicates the range for which the project type has adverse impacts on plan Bay Area targets, whereas the x-axis to the right of the y-axis junction indicates the range for which the project type is supportive of the plan Bay Area targets. Moving from left to right, the project with the least advantageous impact rating is the SR-239 Expressway, with a rating of about -3, although the project had a benefit-cost ratio of about 8. The RS-4 bypass completion rates about -2 and also has a benefit-cost ratio of about 2. The new SR-84/I-680 interchange widening rates about -2, with a benefit-cost ratio of about 4. The new SR-152 alignment rates between -1 and -2 but has a benefit-cost ratio between 3 and 4. The Bay Area Metropolitan Transportation Council express lanes project has a support rating between 0 and 1 but a benefit-cost ratio of 5. The Silicon Valley express lanes project also has support rating of between 0 and 1 but a benefit-cost ratio of about 6. The Marin-Sonoma Narrows (phase 2) project rates at slightly less than 1 and has a benefit-cost rating of about 1 as well. The I-680/SR-4 interchange improvements and widening project also has a rating of slightly less than 1, with a benefit-cost ratio of about 2. The SR-85 auxiliary lanes project rates slightly less than 1, with a benefit-cost ratio closer to 6 than to 7. The Fremont/Union City East-West Connector rates slightly less than 1, with a benefit-cost ratio closer to 7 than to 6. The I-80 auxiliary lanes (Airbase Parkway to I-680) project rates slightly more than 1 in supporting plan targets and has a benefit-cost ratio of 5. The SR-29 high-occupancy vehicle (HOV) lanes and bus rapid transit project has a rating of a little over 2, with a benefit-cost ratio of about 2. The US-101 HOV lanes (Whipple to Cesar Chavez) rates about 3, with a benefit-cost ratio closer to 6 than to 7. The Freeway performance initiative—the project with the greatest project benefits—has a benefit-cost ratio of about 16 and has a support rating of about 4. The intelligent transportation system improvements in Santa Clara and San Mateo Counties has the second greatest benefits. It also has a benefit-cost ratio of about 16 and has a support rating of about 4. The Bay Bridge contraflow lane project has support rating of between 4 and 5, but a benefit-cost ratio of about 2. Treasure Island congestion pricing project also has a support rating of between 4 and 5, but its benefit-cost ratio is nearly 60. Finally, the congestion pricing pilot project has a support rating of about 6, but its benefit-cost ratio is about 45.Return

Figure 12. Line graph. Mobility: average weekday hourly delay. This line graph depicts the average daily vehicle hours of delay by hour of the day for weekdays during 2004, 2005, and 2006. Average daily vehicle hours of delay at 60 mi/h is on the y-axis from 0 to 1,000 h of delay, and hour of the day is on the x-axis from 12 a.m. to 11 p.m. The graph indicates that there is a 0-h delay from midnight through 6 a.m., when delay sharply increases, peaking at 8 a.m. During 2004 and 2005, this peak was at about 500 h of delay. During 2006, that number rose to nearly 700 h of delay. From 8 to 10 a.m., delay drops off as sharply as it increased, averaging around 50 h for the period from 10 a.m. through about 2 p.m. for 2004 and 2005, although that average ranges from 50 to 100 h for the same time period in 2006. At about 2 p.m., delay begins to increase steadily until about 4 p.m., when it begins to increase sharply before peaking at about 5:30 p.m. before declining again. Delay jumps from less than 200 to nearly 600 h in 2004, from about 250 to more than 650 h in 2005, and from 300 to more than 700 h in 2006. Average delay drops off as quickly as it built, dropping to between 200 and 250 h by 7 p.m. for all 3 years. By 8 p.m., delay is largely eliminated and remains flat at 0 h for the remainder of the 24-h period. Return

Figure 13. Line graph. Reliability—buffer index. This line graph depicts travel time by time of day for a 24-h period. The y-axis shows travel time from 0 to 90 min, and the x-axis shows time of day from midnight to midnight. The lines represent average travel time, buffered travel time (95 percent on time), travel time at 60 mi/h, and travel time at 35 mi/h. The lines for both travel time at 60 and 35 mi/h are flat horizontal lines for the entire 24-h period. The average travel time and buffered travel time mirror each other closely, with buffered travel time being slightly greater. Travel time builds from midnight to 3 a.m. to a minor peak, increasing from 34 to about 38 min average travel time (from 35 to about 43 min for buffered travel time) before dropping off to about 30 min at 6 a.m., when travel time begins to increase sharply for the morning peak period. Average travel time builds to peak at about 60 min at about 8:15 a.m. (to about 73 min for buffered travel time) before steadily dropping back down to about 33 min at 1 p.m. Average travel time increases again in the afternoon beginning at about 4 p.m., rising steadily to about 59 min just before 6 p.m. (about 66 min for buffered travel time), when it begins to decline steadily, flattening at about 9 p.m. to an average travel time of about 31 min and a buffered travel time of about 33 min. Return

Figure 14. Illustration. Heat map of bottleneck locations and extents. This heat map indicates the location, time, and extent of bottlenecks on a specific corridor. The map indicates that the Montague Expressway and the Embarcadero/Oregon Expressway experience significant speed reductions from 6 to about 9 a.m., with the Embarcadero Expressway also experiencing an evening bottleneck condition from about 7 to 8 p.m. Ellis experiences significant reductions in speed from 6 a.m. through nearly midday. Hillsdale and 3d on Peninsula and, to a lesser degree, Milbrae also experience morning speed reductions from about 7 to 9 a.m., with Hillsdale and 3d on Peninsula also experiencing significant speed reductions from 7 to about 8 p.m. Return

Figure 15. Screen capture. TEMS. This screen capture of the Traffic Operations System Equipment Management System (TEMS) includes a map, screen navigation functions such as zoom and re-center, options for toggling traffic operations system equipment location layers, equipment status indicator options, and transportation network and boundary layers. Below the map is a tabular list of equipment by ID that indicates maintenance ID, type, county, route, PM, direction, status, functional description, and an option to view more detail about the item. Return

Figure 16. Bar graph. Comparison of corridor segments by delay per vehicle per mile. This graph shows a comparison of corridor segments by delay per vehicle per mile. The x-axis represents delay per vehicle per mile from 0 to 20 min/mi. The y-axis represents various corridor segments in the direction of traffic flow, from bottom to top. The bottom two segments are Bellefield Avenue to Bigelow Boulevard and Bigelow Boulevard to Bouquet Street. There is no delay on either segment during the p.m. peak. Next, Bouquet Street to Craft Avenue experiences about 5 min of delay per vehicle per mile, and Craft Avenue to Moultrie Street experiences about 1 min of delay. The Moultrie Street to Pride Street segment experiences just over 4 min of delay, but Pride Street to Washington Place experiences nearly 16 min of delay per vehicle per mile. The Washington Place to 6th Avenue segment experiences just under 6 min of delay per vehicle per mile. Return

Figure 17. Line graph. Variability in speed caused by incidents as a measure of nonrecurring congestion. This line graph displays speeds on the I-579 Veteran’s Bridge from August 22 to 28, 2005. The x-axis represents time and ranges from midnight on the morning August 22 to 11:59 p.m. on the evening of the August 28 in 6-h increments. The y-axis presents southbound average spot speeds ranging from 25 to 75 mi/h in 5-mi/h increments. In general, the line graph oscillates around 68 mi/h throughout the time span. It contains three distinct downward spikes. The first occurs a little after 6 a.m. on the August 22 and reaches an average speed of approximately 43 mi/h. The second occurs a little after 6 a.m. on the August 23 and reaches a minimum average speed of approximately 27 mi/h. The third occurs a little after 6 p.m. on the August 26 and reaches a minimum average speed of approximately 32 mi/h. A straight horizontal pink line crosses the graph at 50 mi/h and represents the posted speed limit. Text on the graph indicates its data source is Traffic.com. Return

Figure 18. Chart. Policy-level performance targets. This chart indicates policy-level performance targets for economy, environment, and equity. Economic targets include safety, congestion (by 2035 reduce vehicle hours of delay (VHD) per person by 10 percent compared to 2010), and freight reliability. Environmental targets include basic infrastructure, clean air (by 2035 ensure 0 percent population exposure to at-risk levels of transportation-related air pollution), and travel. Equity targets include affordability (by 2035, reduce the share of households in the region spending 50 percent of income on housing and transportation compared to 2000), congestion (by 2035 reduce VHD per person by 10 percent compared to 2010), and access to daily needs. Return

Figure 19. Map. The a.m. peak intersection level of service and travel speed. This figure displays a high-level map of major roadways between Wilmington and Newark, DE. The map displays level of service (LOS) for numerous intersections and small segment travel speeds in and between Wilmington and Newark, with the majority of highlighted spot locations rating an LOS E (defined as 30 to 40 percent of freeflow speed) or F (defined as less than 30 percent of freeflow speed). Multiple segments are rated at an LOS E. Return

Figure 20. Map. Identified congested corridors. This figure displays an aerial map of roadways between Wilmington and Newark, DE. The map highlights congested corridors in and between Wilmington and Newark. Corridors are numbered 1 through 10; however, corridor names are not provided on the map. No legend is provided. Return

Figure 21. Map. The p.m. peak travel speed changes, 2004–2011. This high-level map of roadways between Wilmington and Newark, DE, highlights roadway segments in the region that have experienced extreme evening peak travel speed changes. Several segments are colored red, indicating they experience a 25 percent or greater decrease in travel speed, while green segments indicate they experience a 25 percent or greater increase in travel speed. Return

Figure 22. Map. PAG relative congestion. This Pima Association of Governments (PAG) map depicts two-way weekday traffic counts on major roadways in Oro Valley adjacent to Catalina State Park. The map’s road lines are different thicknesses, representing different volumes of traffic on each road. A small colored shape is placed on top of each major road segment, with each shape representing a different year from 2006 to 2011. Most are for 2010, but there are no shapes for 2011 or 2006. Each shape also contains a number representing the traffic volume on the road segment. Shapes on the map contain numbers ranging from 1 to 34 representing traffic count in thousands. The legend also notes that traffic volumes are derived from PAG, city, county, and State traffic counting devices. Seasonal variation is not factored into the volumes. Return