Technical Memorandum

DATE:   December 19, 2019

TO:         Boston Region Metropolitan Planning Organization

FROM:   Ryan Hicks and Seth Asante, Central Transportation Planning Staff

RE: New and Emerging Metrics for Roadway Usage

1          Introduction

The Boston region contains a robust transportation network that encompasses diverse modes of transportation, including vehicular, truck, public transportation, bicycle, and pedestrian movement. Yet, these different modes of transportation are monitored separately as transportation performance monitoring has traditionally focused on moving vehicles rather than people. To better monitor multimodal travel as it relates to the mobility of individual travelers, the Boston Region Metropolitan Planning Organization (MPO) supported the New and Emerging Metrics for Roadway Usage study through its federal fiscal year (FFY) 2019 Unified Planning Work Program (UPWP).

The objectives of this study were explicitly stated in the UPWP work program. The objectives were to 1)  determine performance measures that can assess multiple modes and quantify the mobility of motorists, transit riders, bicyclists, and pedestrians rather than vehicles, and 2) determine a plan for the selected performance measures to be considered for the Congestion Management Process (CMP), the Long-Range Transportation Plan (LRTP), and other MPO activities. These multimodal monitoring criteria may significantly benefit the programs listed above, but it is not obligatory for any agency or program to adopt the recommended criteria. Though not required, the recommendations from this study will be useful for educating planners, engineers, and the public about multimodal performance monitoring.

The first section of this memorandum focuses on a literature review of studies that focused on multimodal transportation performance monitoring. The memorandum further discusses the selection process for performance metrics based on the goals and objectives that are in the Boston Region MPO’s LRTP, Destination 2040. Then there is a summary of a test analysis that applied selected performance metrics to conditions at two locations on the transportation network. The memorandum concludes by presenting the selected performance metrics and the adjusted thresholds based on the previously mentioned analysis.

2          Literature Review

2.1      Background

Over the last few years, there has been ample research on developing nuanced performance metrics that are multimodal and inclusive. As a result, many transportation organizations are evolving traditional level of service (LOS) standards into multimodal LOS standards. Thus, it was imperative for the MPO staff to review existing multimodal studies and guidelines before creating a unique multimodal performance monitoring method for voluntary use in MPO activities, such as corridor and intersection studies. Elements from other MPO studies, such as those that resulted in the development of the Bicycle Report Card and the Pedestrian Report Card Assessment tool, were incorporated into this study.   

An ideal multimodal performance monitoring method should include performance metrics and land use elements, and have minimal influence from transportation demand models. For multimodal performance monitoring, it is ideal to have data that represent real conditions on the transportation network rather than modeled data as future projections can be unpredictable due to unforeseen changes to planned transportation developments. This section summarizes several studies that relate to the topic of multimodal performance monitoring and their strengths and weakness.

2.2      Victoria Transport Policy Institute

The How Should We Measure Traffic Congestion? study was developed by the Victoria Transport Policy Institute (VTPI), an independent transportation planning organization based in Victoria, British Columbia.1 The goal of VTPI is to determine new ways to fix transportation-related problems. This study emphasized the interesting point that intensity-related performance measures are good for determining short-term planning but ignore congestion exposure, which is defined as “the amount people must drive under urban-peak conditions.”

VTPI was cognizant of newer technologies, such as app-based ride-sharing and autonomous vehicles, which might increase congestion by adding to the number of vehicles on the transportation network. This study analyzed the costs of congestion and presented some alternatives to roadway LOS metrics, such as the following:

• multimodal level of service
• quality of service
• trip generation and vehicle travel and fuel consumption models
• multimodal accessibility modeling

This study relates to several goals of the New and Emerging Metrics for Roadway Usage study, as the VTPI study encouraged the improvement of overall accessibility and transport system efficiency rather than maximizing vehicle speeds.

Strengths

This study searched for performance measures that were comprehensive and multimodal.

Weaknesses

This study relied heavily on model-based planning.

2.3      National Cooperative Highway Research Program

The National Cooperative Highway Research Program’s Field Test Results of the Multimodal Level of Service Analysis for Urban Streets study relied upon a few principles, such as measuring congestion exposure rather than congestion intensity as intensity indicates how extreme congestion is at a specific time.2 Included in this study were cases that analyzed and assessed performance metrics at several locations in the United States. The main goal of this study was “to develop and test a framework and enhanced methods for determining levels of service for automobile, transit, bicycle, and pedestrian modes on urban streets, in particular with respect to the interaction among the modes.” Examples of the performance metrics tested include pavement condition and driveway conflicts per mile.

Strengths

• The study made a concerted effort to incorporate several transportation modes into the calculations.
• Refinement of the model takes place after the assessment has concluded, which improves the accuracy of the model.

Weaknesses

This study relied on models, which might not represent real conditions on a transportation facility. To improve accuracy, model calibration is necessary to reflect local conditions.

2.4      Bellevue Transportation Commission

The Bellevue Transportation Commission of the City of Bellevue, Washington, proposed a transportation monitoring system in its MMLOS Metrics, Standards and Guidelines report (2017-18) that accommodates all travelers and all trips, regardless of mode.3 The study was initiated in response to Washington State’s Growth Management Act of 1990, which requires local governments to observe LOS on city-owned roadways and transit routes. Strategies and policy revisions are then recommended based on those LOS calculations. The study report states that implementing a multimodal LOS monitoring system can progress the City towards a comprehensive citywide multimodal transportation system. The study recommended using the following LOS metrics for the vehicle, pedestrian, bicycle and transit modes:

• The vehicle LOS measures are based on the Highway Capacity Manual (HCM) standards.
• The pedestrian LOS measures focus on buffers and landscaping, which can affect pedestrian comfort. Crossing opportunities are also analyzed.
• The bicycle LOS measure analyzes the level of stress for bicyclists using bicycle facilities, such as bike lanes and bike paths.
• The transit LOS measures examine factors that affect ridership or the experience of the transit rider.

Strengths

• This study evaluated all modes equally through the recommended performance metrics.
• The recommendations from this study enabled the City of Bellevue to make changes to its transportation policies to evaluate transportation corridors and monitor transportation performance using performance targets.

Weaknesses

The metrics for vehicle and transit modes may not be sufficient for large urban areas, which must focus on air quality and climate change issues and are aiming to reduce vehicle-miles traveled (VMT) while increasing transit and nonmotorized mode shares.

2.5      City of Ottawa

The report titled Multimodal Level of Service (MMLOS) Guidelines was completed for the City of Ottawa, Ontario, by the IBI Group.4 This study examined a way to measure congested conditions without exclusively favoring roadways. This study analyzed other modes of transportation, such as the pedestrian, bicycle, public transportation, and truck modes. As a result, the City of Ottawa created the MMLOS tool in 2015. This tool emphasizes evaluating tradeoffs for improving one mode over another mode.

The metrics selected for this tool are measured at both signalized roadway intersections and roadway segments except for vehicular LOS metrics, which can only be measured at intersections. The metrics are as follows:

• The pedestrian LOS criterion focuses on pedestrian comfort, safety, and convenience.
  
  
• The bicycle LOS criterion focuses on level of traffic stress (LTS) experienced by bicyclists at specific roadway locations.
  
  
• The transit LOS criterion focuses on the transit mode share of a corridor by analyzing transit travel time and transit priority afforded to transit vehicles.
  
  
• The truck LOS criterion focuses on the percentage of trucks and buses in a corridor. Buses and trucks are monitored together because they are considered to be heavy vehicles.
  
  
• The vehicular LOS criterion focuses on the functionality of intersections according to the Transportation Impact Assessment Guidelines. 5

Strengths

• Trade-off evaluations are very helpful for determining the best way to solve congestion in a corridor.
• This study uses performance metrics to evaluate all modes of travel.

Weaknesses

• This method of performance monitoring requires a large, varied data collection effort.
• The study emphasized measuring LOS but fell short on measuring quality of service (as measured by mode shares, VMT, and the like). The monitoring of quality of service is useful for tracking policy goals.

2.6      California Senate Bill 743

California Senate Bill 743 was enacted in 2013 and will mandate on July 1, 2020, changes to the guidelines for implementing the California Environmental Quality Act (CEQA) in regards to the analysis of transportation impacts.6 This bill imposes changes to the way California municipalities will measure the environmental effects of projects by focusing on the relationship between transportation and land use.

This bill requires municipalities to eliminate LOS and vehicle delay measures and begin monitoring VMT to gauge transportation impacts. A reason for this change is that the previous LOS requirements of the CEQA encouraged sprawl in certain locations while discouraging infill development closer to jobs, transit and walkable areas.7

As stated in the Shifting Gears in Transportation Analysis webinar, issues with LOS as a measure of transportation impact are as follows:8

• “Punishes last-in [developments constructed near the central business district] inhibits infill, pushes development outward”
  
  
• “Solves [or reduces] local congestion, exacerbates regional congestion”
  
  
• “Inhibits transit”
  
  
• “Inhibits active transport”
  
  
• “Measures mobility, not access; shows failure when we [local and regional governments] succeed [for example, sprawl can increase mobility of travelers, but it reduces the access to locations to the same travelers]”
  
  
• “Measures mobility poorly; fails to optimize network even for automobiles”
  
  
• “Forces more road construction than we [local and regional governments] can afford to maintain”
  
  
• “Hard to calculate and inaccurate”

Strengths

• The bill promotes a more multimodal performance monitoring approach for local communities.
  
  
• The bill helps municipalities monitor both land use and transportation.
  
  
• Projects that promote sprawl will have to account for the extra traffic that will occur as a result.
  
  
• This bill encourages local planners and officials to consider creating their own performance measures.

Weaknesses

It is unknown if VMT will be an adequate metric to monitor congestion as new technologies such as autonomous vehicles emerge.

2.7      Guide to Sustainable Transportation Performance Measures

The Guide to Sustainable Transportation Performance Measures, published by the US Environmental Protection Agency in 2011, benefited from several interviews and information exchanges with staff from several MPOs, the Federal Highway Administration, and the Federal Transit Administration.9 The document describes opportunities to incorporate environmental, economic, and social sustainability into transportation decision-making through the use of performance measures. Among the several performance measures described in the guide are transit accessibility, transportation affordability, and VMT per capita. 

Transit accessibility metrics indicate the relative convenience of transit as a mode choice. Transit accessibility can be measured in terms of the distance people must travel to transit stops or travel time on transit. These metrics typically emphasize the availability of transit where people live, where people work, and on commuter routes.

Transportation affordability metrics reflect the ability of transportation system users to pay for transportation. Since an affordable transportation system is one that takes a smaller share of a user’s total income, typical metrics include annual transportation costs compared to annual income and the proportion of income spent on transportation by people in various income groups.

VMT per capita metrics show the amount of vehicle activity per population in a region. Reducing VMT has been associated with better air quality, congestion reduction, and fewer vehicular crashes.

Strengths:

• The guidebook provides examples of best practices in sustainable transportation performance measurement that are being applied across the country.
  
  
• Data for most of the performance metrics are available to most MPOs or can be obtained from other planning agencies.
  
  
• The recommended monitoring methods from the guidebook incorporate environmental, economic, and social (equity) sustainability into transportation decision-making.

Weaknesses:

The method is sourced from various MPO regions; therefore, it is not customized for large MPO regions with multimodal transportation systems such as the Boston region.

2.8      SUMMARY

The features of the six studies that are most pertinent to multimodal transportation planning are summarized in Table 1. Based on the research conducted, the two studies that served as models for the New and Emerging Metrics for Roadway Usage study were the MMLOS studies in Bellevue, Washington, and Ottawa, Ontario. The guidelines from these studies provide performance metrics for multiple modes, incorporate a land use component, and feature very little influence from a model.

A general trend is that many transportation agencies are minimizing or eliminating the use of automobile LOS and promoting the use of VMT. Most of these studies agree that the best way to accurately measure congestion across multiple modes is to determine ways to measure the movement of people rather than the movement of vehicles.

Table 1
Study Comparisons

LOS = level of service. MMLOS = multimodal level of service. NCHRP = National Cooperative Highway Research Program. VMT = vehicle-miles traveled.

Source: Central Transportation Planning Staff.

3          Destination 2040 Goals and Objectives

For performance monitoring to be effective, goals and objectives must be well-defined to enable the selected metrics to gauge progress towards improving the transportation network. The MPO defined six goal areas and associated objectives in its LRTP, which was completed in August 2019:

• Safety
• System Preservation and Modernization
• Capacity Management and Mobility
• Economic Vitality
• Clean Air/Sustainable Communities
• Transportation Equity

Three of those goal areas—Safety, Capacity Management and Mobility, and System Preservation—are relevant to this study and the selection of performance metrics. Each of these goal areas pertain to moving people safely and effectively across multiple modes. Monitoring the mobility of individual travelers rather than vehicles will show a more accurate depiction of congestion on the transportation network. Person capacities for transportation facilities often are higher when vehicle occupancies are higher. These three goals and 11 associated objectives are discussed further below.

Safety

The MPO’s safety goal focuses on the safety of travelers for all modes. Oftentimes, this goal involves analyzing Highway Safety Improvement Program (HSIP) crash cluster locations. The LRTP states that facilities to improve safety for bicyclists and pedestrians are required to improve safety at high crash locations as bicyclists and pedestrians are involved in a growing share of crashes.

The safety objectives include the following:

• Reduce number and severity of crashes  and safety incidents for all modes
• Reduce serious injuries and fatalities from transportation
• Make investments and support initiatives that help protect transportation customers, employees, and the public from safety and security threats 

Capacity Management and Mobility

The Capacity Management and Mobility goal focuses on the movement of people and goods throughout the transportation network and the connectivity of the transportation network. The goal emphasizes the need to ensure that transportation infrastructure meets the Americans with Disabilities Act (ADA) standards. This goal is consistent with the goals that were presented in the studies from the literature review, with the focus on multiple modes rather than exclusively automobile travel.

The Capacity Management and Mobility objectives include the following:

• Improve reliability of transit
• Support implementation roadway management and operations strategies to improve travel reliability, mitigate congestion, and support non-single-occupant vehicle travel options.  
• Fund improvements to bicycle/pedestrian networks aimed at creating a connected network of bicycle and accessible sidewalk facilities (at both regional and neighborhood scale) by expanding existing facilities and closing gaps
• Support strategies to better manage automobile and bicycle parking capacity and usage at transit stations
• Increase percentage of population and places of employment within one-quarter mile access to transit stops
• Improve access to and accessibility of all modes, especially transit and active modes
• Eliminate bottlenecks on the freight network and improve freight reliability

System Preservation

The System Preservation goal focuses on maintaining the existing transportation network. The areas of focus include, but are not limited to, infrastructure located along roadways and at intersections, including sidewalks and pedestrian signals.

The System Preservation objectives include the following: Improve existing pedestrian and bicycling infrastructure.

4          Data Availablity

An effective performance monitoring program begins with data collection. Transportation data can be obtained by manual collection or by receiving or purchasing data from a third party. Table 2 shows the prospective performance metrics and data sources that were evaluated in the studies described in Section 2 of this memorandum. The MPO staff already has access to some of these datasets and others would be easily obtained. Some other datasets would be difficult to obtain or contain incomplete data. Some metrics can currently only be collected by doing manual fieldwork, which can be expensive and time-consuming. Only the metrics that rely on data sources readily available to the MPO staff were selected for this study.

Table 2
Data Sources and Prospective Performance Metrics

* INRIX is a private company that collects roadway travel times and origin-destination data for most roadways that are collectors, arterials, limited-access roadways or freeways.

^† Data for other local regional tranist authorities and organizations—such as the Brockton Area Transit Authority, Cape Ann Transportation Authority, and MetroWest Regional Transit Authority—may be substituted for local datasets, if necessary.

^‡ The National Performance Management Research Data Set (NPMRDS) includes archived INRIX vehicle probe data and freight probe data. The American Transportation Research Institute provides the freight probe data that is included in the NPMRDS. RITIS provides the NPMRDS and incident data to the Boston Region MPO through its web portal.

ADA = Americans with Disabilities Act. MassDOT = Massachusetts Department of Transportation. MPO = Metropolitan Planning Organization.  

Source: Central Transportation Planning Staff.

5          Selection of Performance metrics

Prior to implementing a new performance monitoring program, it is important to conduct outreach and receive feedback on the proposed performance metrics from stakeholders. This study included two outreach efforts involving an online survey and interviews. An online survey is best for receiving quantitative feedback from a large number of respondents and interviews are ideal for collecting in-depth feedback from stakeholders.

5.1      Survey

The online survey was distributed between April 23, 2019, and May 10, 2019, to approximately 55 professionals who work in the field of transportation planning and are involved with performance monitoring throughout the New England region. Overall, 17 survey responses were received. A copy of the final survey is located in Appendix C. The survey asked respondents about their

• general background (affiliation);
• preferred mode of travel;
• opinion about the definitions of potential performance metrics that were derived from the studies presented in the literature review and various MPO activities;
• rankings of potential performance metrics by travel mode;
• opinion about multimodal transportation monitoring in areas with certain demographic characteristics; and
• advice about any metrics that should be added or changed.

For an extensive analysis of the survey results, refer to Appendix D. The survey results indicated that transit and pedestrian modes rank the highest in regards to preferred mode of transportation. Several of the respondents stated that they would like to see performance metrics that show connections between different modes. Measuring automobiles and travel speeds are very polarizing as many of the respondents reported that they were strongly for or against including automobile speeds.

The key findings of the survey were as follows:

• The curb radius and pavement condition metrics received negative feedback. The suitability of these metrics can vary based on the geometric design of an intersection or roadway segment.
  
  
• Monitoring of vehicle travel speeds is not perceived positively for multimodal transportation performance monitoring.
  
  
• Participants suggested adding performance metrics for intermodal connectivity and viability of weather by mode (comfort of mode of travel during inclement weather). Intermodal connectivity can be measured in part by the presence of bicycle racks and safe crossings at transit stops. Viability of weather is factored in the form of comfort, in regards to various bicycle, pedestrian, and transit metrics.

5.2      Interviews

The MPO staff interviewed several transportation planning professionals in the Boston region to get more extensive feedback on the potential performance metrics. The interviewees recommended that the study focus on both the comfort and movement of people and that five or six metrics per travel mode should be used. The interviewees recommended some modifications to the proposed metrics and suggested some new metrics.

The interviewees recommended including new metrics such as low-stress bicycling and network connectivity, number of bike lane discontinuities, crossings at transit stops, and number of pedestrian interruptions. Crossings at transit stops was selected as a metric for this study. The other recommended metrics were not selected, but features from the metrics were interwoven into other metrics that were selected.

The interviewees also recommended changing the crossing opportunities metric to safe crossing opportunities and adding criteria to evaluate if a crosswalk is safe for a pedestrian to cross in the time provided by the traffic signal. Interviewees also recommended that the transit time index measure should be based on a free-flow travel or baseline travel time, rather than the transit schedule.

5.3      Selected Performance Metrics

Based on the findings from the literature review and the survey results, MPO staff recommend that 24 performance metrics to be used for evaluating multimodal transportation facilities. Table 3 lists the recommended performance metrics according to the modes they measure and the Boston Region MPO goals to which the metrics relate.

Table 3
Selected Performance Metrics

* This metric was previously used by Boston Region MPO.

^† Sidewalk capacity for pedestrians was not directly measured in this study.

^‡  For the duration of congestion/congested time metric, the definition of congestion on arterials is the total time when travel speeds are below 19 miles per hour.

Source: Central Transportation Planning Staff.

6          Performance metric definitions

This section provides a brief overview of the performance metrics that were selected for the multimodal performance monitoring criteria. For a detailed description of each performance metric and the data required for monitoring, please refer to Table A.1 in Appendix A.

6.1      Bicycle Metrics10

Bicycle Crashes

The bicycle crashes performance metric analyzes the safety of a corridor, based on the number and severity of bicycle crashes. The crashes will be assessed based on the bicycle EDPO score at the intersections in the corridor.11 This metric is different from the absence of bicycle crash metric that was presented in the Boston Region MPO’s bicycle level of service metric study,12 which rates bicycle safety by the presence of HSIP clusters. 

Bicycle Facility Continuity (Bicycle Facility Presence)

The bicycle facility continuity metric examines the length of a bicycle facility (such as a bicycle lane) compared to the roadway segment where the bicycle facility is located.

Level of Traffic Stress

The metric for level of traffic stress experienced by bicyclists is based on vehicular travel speeds, vehicular volumes, and the presence of buffers between vehicles and bicyclists. There are various degrees of stress that bicyclists can experience on a roadway segment or corridor, which would determine the recommended experience needed for a bicyclist to traverse through a roadway segment or corridor. Detailed tables that demonstrate the process for calculating level of stress are located in Appendix B.13

Bicycle Rack Presence

The bicycle rack presence metric indicates if a corridor has bicycle parking available nearby. This may include bicycle racks along a street, bicycle racks located near a transit station or bus stop, or bicycle racks that are located on nearby private property where any bicyclist is permitted to park a bike. Racks located at a transit or bus stops give the additional benefit of providing a connection to other modes for travelers. 

Proximity to Bike Network

The proximity to bike network performance metric evaluates ways that a roadway segment serves as a connection to a bicycle route. Roadway segments within one-quarter mile of a bicycle facility that provide bicycle accommodations that separate bicyclists from mixed traffic are ideal.

6.2      Pedestrian Metrics14

Safe Crossing Opportunities/Safe Crosswalks per Mile15

The safe crossing opportunities performance metric reflects the number of crosswalks that are present along roadway segments. This metric is reported as the number of crosswalks per linear mile.

Safe crossing opportunities per mile = number of safe crosswalks along a roadway segment/length of roadway segment in miles

Sidewalk Presence and Condition

The sidewalk presence performance metric indicates whether sidewalks are present along a roadway segment or at an intersection and are in good condition.16 Sidewalks that are valid for evaluation are those that meet American with Disabilities Act (ADA) standards. This metric is measured by each direction of travel.

Sidewalk presence and condition (calculated for each individual direction of travel) = total length of sidewalks in good condition/total length of roadway

Pedestrian Crashes

The pedestrian crashes performance metric documents areas where pedestrian crashes are common. This performance metric will be assessed based on the EDPO in the corridor. The EDPO score is presented in a per mile basis for the entire corridor.

Vehicle-Pedestrian Buffer

The vehicle-pedestrian buffer measures the total distance between vehicular traffic and pedestrian traffic. The vehicle-pedestrian buffer includes any infrastructure that is present between a vehicle travel lane and an adjacent sidewalk or walkway. A buffer helps reduce vehicle-pedestrian traffic incidents, which often result in bodily injuries or fatalities.

6.3      Transit Metrics17

Transit Time Index

The transit time index compares the average travel time of a transit vehicle to the tenth percentile daily travel time of the daily bus run. This measure can be used to calculate delay along a transit route.

Transit time index = average travel time/tenth percentile daily travel time

Level of Transit Time Reliability

The level of transit time reliability metric measures the variation of the travel time for a transit route during a typical weekday. This metric indicates if there is variability or consistency in travel times on a route from day to day. MBTA timepoint crossing summary data will be used to measure this metric.

Level of transit time reliability = 80th percentile travel time for transit vehicle/50th percentile travel time

Person Hours of Delay per Bus Trip

The person hours of delay per bus trip metric combines ridership numbers with the travel time delay of transit vehicles. The delay for each run can be multiplied by the average ridership. The hours of delay can be calculated for the peak period, entire day, or entire year.

Person hours of delay = transit vehicle delay * average number of people on transit

Vehicle Delay per Bus Run

The vehicle delay per bus run metric shows the average vehicle delay per trip for a bus route regardless of service frequency.

Bus run delay = (average travel time for bus run + departure delay time) – free-flow or baseline travel time

Load Factor/Passenger Crowding

Passenger crowding is measured as the ratio of the number of passengers on a vehicle at the maximum load point to the number of seats on the vehicle.

Passenger crowding = number of passengers on the vehicle/number of seats on the vehicle

Safe Crossing Opportunities at Transit Stops

The safe crossing opportunities at transit stops metric analyzes the percentage of transit or bus stops in a corridor that have safe crossings nearby.

6.4      Truck Metrics18

Truck Travel Time Reliability Index

The truck travel time reliability index was introduced by the Federal Highway Administration and is calculated using the National Performance Management Research Dataset. This metric can be used to identify predictable bottlenecks on individual roadway segments. Values on this index are calculated by dividing the 95th percentile travel time by the 50th percentile travel time.

Truck travel time reliability index = 95th percentile travel time/50th percentile travel time

Percentage of Truck Traffic

The percentage of truck traffic metric shows the percentage of vehicles on a roadway that are trucks. This performance metric is a useful tool for prioritizing transportation projects. Corridors with a high percentage of truck traffic have different needs from corridors that have little truck traffic. Therefore, if two projects have equal evaluation scores, the project that is located on the corridor that has a higher percentage of trucks may be prioritized higher with respect to freight needs. This metric is not rated as good, average, or poor. Instead, this metric is rated as low truck traffic, medium truck traffic, and high truck traffic.

Buffer Time per Trip/Total Hours of Daily Truck Buffer Time

The buffer time per trip and total hours of daily truck buffer time metrics indicate the amount of contingency time that freight providers would need to consider to ensure that a truck trip is completed on time 95 percent of the time. These metrics can either be calculated by analyzing an individual truck trip (buffer time per trip) or the total truck trips (total hours of daily truck buffer time) that are made daily. Finding the total daily truck buffer time would require data on daily truck volumes.

Buffer time per trip (minutes) = 95th percentile travel time – average travel time

Total hours of daily truck buffer time (hours) = (95th percentile travel time – average travel time) * truck volumes

6.5      Vehicle Metrics19

Duration of Congestion/Congested Time

Congested time is the average number of minutes that drivers experience congested conditions (speeds below 19 miles per hour [mph] on arterials), during a peak period. Congested time is measured in minutes per peak period hour.

Congested time (minutes) = (number of minutes with speeds below 19 mph/total number of minutes in sample) * number of minutes in peak period

Travel Time Index

The travel time index compares travel conditions during the peak period to travel conditions during free-flow periods. The travel time index is the ratio of peak period time to free-flow time.

Travel time index = average travel time/free-flow travel time

Vehicle-Miles Traveled

Vehicle-miles traveled (VMT) are the total number of miles that every vehicle travels through a roadway segment, corridor, or region within a specified period of time. This metric is becoming a basis for measuring transportation patterns, as evidenced by California Senate Bill 743. Additionally, many transportation departments across the United States are switching from LOS-based metrics to VMT-based metrics. The reason for this change is that it is desirable to determine if a proposed project will result in an increase in VMT in the surrounding area.

Vehicle-miles traveled = segment length X vehicle volumes

Average Vehicle Delay per Mile

The average vehicle delay per mile metric shows the amount of delay that a vehicle would experience when traveling through a corridor at a designated time.

Average vehicle delay per mile (seconds) = (average travel time for monitoring period – free-flow or baseline travel time)/length of corridor

6.6      Multimodal Metrics20

Roadway Lane Density

Roadway lanes are most effective when the throughput of people is maximized. Data for the roadway lane density metric will be based on observations of travelers passing through a corridor during a specified period of time. The type of vehicles, vehicle volumes, and vehicle occupancies will be recorded. Reducing the percentage of single-occupancy vehicles and increasing the percentage of vehicles that have high occupancies, such as buses, would help increase roadway lane density. Bicyclists and pedestrians are not included in this metric.

Roadway lane density = (vehicle volumes for one hour period * occupancy counts for one hour period)/number of lanes

Person Throughput

Person throughput is a time-based metric that indicates the number of people attempting to enter a segment or corridor during a specified monitoring period. All parallel transportation facilities, including sidewalks and bicycle lanes, are included in this metric. This metric reflects the number of people who travel in a corridor by walking, biking, taking the bus, and driving in an automobile. A higher person throughput indicates that a transportation facility is moving more people during a specified time. The thresholds for this metric can vary depending on the evaluator’s choice (rural versus urban standards, for example).  

Peak hour person throughput = vehicle, bicycle, and pedestrian volumes for one-hour period * occupancy counts for one-hour period

7 Corridor Selection for metrics testing

Only arterial corridors monitored on the CMP network and identified as an LRTP priority corridor or a subregional priority roadway were considered for evaluating the potential metrics for calibration. The two corridors selected for the testing were as follows:

• Route 16 in Medford between the Mystic River and the Everett city line
• Route 9 in Brookline between the Newton city line and Washington Street

These corridors are between one and five miles long, which is a corridor length that can accommodate travelers of all modes. The selection process ruled out corridors where projects funded through the MPO’s Transportation Improvement Program (TIP) have recently occurred and corridors where construction is currently underway. Oftentimes, construction activities distort performance monitoring for all modes.

8          Performance metrics test and Calibration

8.1      Procedure

The actual conditions on the two corridors were analyzed for the AM peak period (6:30 AM to 9:30 AM) to determine the thresholds that should be applied to the performance metrics. Additionally, MPO staff surveyed AM peak hour (8:00 AM to 9:00 AM) conditions on-site. Performance metrics were rated based on the data collected at the corridor and from various sources. The performance metrics for each corridor ranked as excellent, average, or poor, based on the actual values of the performance metrics versus the thresholds.

The thresholds for most metrics were set based on how they were used in previous programs, such as the CMP, bicycle and pedestrian activities, and studies done by other organizations. The thresholds for new metrics were determined by staff judgements that were based on evaluations from the test runs compared to the observation of real-time conditions. These thresholds are deemed tentative and are subject to change based on further input. Additionally, evaluators can change the thresholds to cater to their study, if desired. Table 4 presents the thresholds for each individual performance metric and the source of the thresholds. Please see Table A.1 in Appendix A for the threshold sources for each individual metric.

Table 4
Thresholds for Performance Metrics

Note: The percentage of truck traffic performance metric is not rated on this scale. This metric is rated as low truck traffic for corridors with less than four percent truck traffic, medium truck traffic for corridors with four percent to six percent truck traffic, and high truck traffic for corridors with more than six percent truck traffic. 

EPDO=Equivalent Property Damage Only Index. N/A= not applicable.

Source: Central Transportation Planning Staff.

8.2      Evaluation Results

Travel lanes for both directions on Route 9 (in Brookline between the Newton city line and Washington Street) and Route 16 (in Medford between the Mystic River and the Everett city line) were evaluated during the AM period of 6:30 AM to 9:30 AM and during the peak hour of 8:00 AM to 9:00 AM on-site to determine how efficiently travelers moved through the corridor using various modes of transportation. Table 5 shows the performance metric values for these sections of Route 9 and Route 16. Table 5 also shows the performance metric ratings, which indicate mobility on the corridors.

Route 9

Travel on both directions of Route 9 is not very suitable for bicyclists or pedestrians. On Route 9 eastbound, however, the bike path surrounding the Brookline Reservoir provides a partial bicycle and pedestrian connection. Transit on Route 9 rates well on comfort-based performance metrics but poorly on mobility-based measures, indicating that transit riders may experience delays in this corridor. There is a moderate percentage of truck traffic, but truckers require a significant buffer time to travel this corridor. Travelers in personal vehicles experience significant delays in both directions, as evidenced by congested time, travel time index and vehicle delay. Person throughput is typical for a multimodal corridor.

Route 16

There are facilities and connections to trails for bicyclists on Route 16, but bicyclists experience a high level of stress because they must share the roadway with vehicular traffic. This corridor is also not pedestrian friendly, as there are safety and comfort concerns. Buses on Route 16 eastbound experience significant delays during the peak periods. There is significant truck traffic on Route 16, and truckers need to budget significant amounts of buffer time when traveling this route. There is a moderate amount of vehicular delay on Route 16 during the AM peak period. Person throughput is high on Route 16 as there are several buses that travel through this corridor, which leads to Wellington Station.

Table 5
Performance Metrics Data
Route 9 in Brookline between Newton City Line and Washington Street and Route 16 in Medford between the Mystic River and Everett City Line

Blue = Good  (1*)     

Black = Average (and low, medium, or high truck traffic) (2*)

Red = Poor (3*)  

EPDO=Equivalent Property Damage Only.

Source=Central Transportation Planning Staff.

9          Study Findings

9.1      Findings

A good multimodal monitoring program should include multiple performance metrics. It is best to use four to six performance metrics to measure each mode. This allows for an evaluation of both comfort and mobility factors for each mode operating in a corridor. Each individual metric should be scored separately to identify specific deficiencies within a corridor and to determine strategies to better accommodate multimodal travel. Additionally, even though metrics such as intermodal connectivity and viability of weather were recommended from survey respondents, certain metrics were not used for this study while elements of those metrics were incorporated into other selected metrics.

Lane density and person throughput are very important metrics for factoring vehicle occupancies for different modes. These measures penalize corridors that have a high number of single-occupancy vehicles. It is very important to increase the number of people traveling through a corridor without increasing the number of vehicles. In addition to mobility, traveler comfort is very important to monitor, especially with regard to pedestrians, bicyclists, and transit riders. If traveler comfort is poor for pedestrians and bicyclists, new travelers many not be attracted to these modes and there could be an increase in single-occupant vehicle travel.

Obtaining real-time data on-site is more useful for multimodal performance monitoring than modeling data.  Modeled data is problematic because it focuses on volume versus capacity rather than on the movement of people. Also, traveler comfort is often not monitored with modeled data. 

9.2      Recommendations

Though not required, MPO staff recommend that the Boston Region MPO use the multimodal performance monitoring criteria presented in this memorandum for assisting with various studies. These criteria can be used in corridor studies and CMP activities, and some criteria can also be considered for use in the LRTP and the TIP. The use of the selected performance metrics would help fulfill the MPO’s goals of improving safety, capacity management and mobility, and preservation of the transportation network in the Boston region. They would also be excellent tools for communities to use to rate their transportation facilities and would help public officials prioritize and secure funding for transportation projects that facilitate the movement of people rather than the movement of vehicles.

9.3      Next Steps

The next step is to refine the performance measures based on feedback from stakeholders. In addition, the evaluation criteria will need to be promoted through outreach to regional planners, engineers, and the public, so that these stakeholders can consider incorporating this tool into their planning processes, if desired. The criteria that are derived from this study will be used in some upcoming Boston MPO corridor studies. Also, the MPO should analyze this process to determine if the evaluation process is suitable for other MPO practices and planning efforts, such as the CMP, LRTP, and the TIP.

RH/rh

Appendices