Multimodal System Performance Measures Research and Application: Innovation and Research Plan

Executive Summary

Background and Purpose

The Intermodal Surface Transportation Act (ISTEA) of 1991 shifted the focus of transportation policy from building the national highway network to integrating multimodal transportation systems. The Moving Ahead for Progress in the 21^st Century (MAP-21) Act of 2012, continuing the policy goals of ISTEA, ushered in a performance-based approach to the Federal Aid Highway program. During the development of the MAP-21 system performance measures (the third performance management rule, sometimes referred to as PM3 or 23 CFR 490.500-490.800), the Federal Highway Administration (FHWA) received thousands of comments, including some asking for multimodal measures that quantify person movements across all modes rather than vehicle movements. Based on those comments, FHWA committed to conduct additional research on multimodal measures and the data needed/used to support them and to report those results. This report is considered the results of this additional research.

The goal of this inquiry, the Multimodal System Performance Measures Research and Application study, was to identify multimodal system performance measures that assess the actual performance of all modes, including light and heavy vehicles, bus and light rail, and non-motorized (i.e., bicycle and pedestrian) travel from a user perspective. The research focused on identifying existing and potential multimodal data sources necessary to develop a true multimodal system performance measure (or suite of measures) and then pilot test what was proposed. This Innovation and Research plan lays out potential next steps in acquiring the data necessary to calculate the proposed multimodal measures and identifies necessary additional research.

The work effort for the Multimodal System Performance Research and Application study was organized around the following work tasks:

1. Conducting a literature review.
2. Defining an "ideal" multimodal system performance measure. Note: the term "ideal" is in quotation marks as a recognition that it would be difficult to identify a truly ideal measure, but that the goal of this task would be to get as close as possible to "ideal."
3. Determining the gaps between data required by the identified "ideal" measure and data currently available.
4. Developing surrogate measures based on available data.
5. Identifying and testing the surrogate measures in three pilot locations.
6. Preparing an innovation and research plan based on the findings of the research.

This research effort relates to but differs from other multimodal measurement initiatives, such as multimodal accessibility and connectivity. This research focuses on the performance of the multimodal system in terms of the productivity of actual person trips made across the multimodal network. Accessibility focuses on the ability to reach destinations while connectivity quantifies the seamlessness of travel across modes.

Complete Trip Perspective and Framework

Many currently used transportation performance measures are carryovers from the pre-ISTEA policy era that focused on individual modes rather than the coordination of modes. Roadway performance continues to rely on congestion-based measures. Some use the Highway Capacity Manual,¹ others use travel time data to determine delay and reliability. Facility based transit measures, such as passengers per route mile, focus on transit productivity and efficiency. Most bicycle and pedestrian measures focus on facility conditions, such as traffic volumes on adjacent streets. Commonly used system measures, such as vehicle miles traveled (VMT) and person hours traveled (PHT), provide feedback on the efficiency of travel across a modal network and, in some cases, across the multimodal system. Such measures reflect the systems perspective sought by this research, but do not provide a user perspective, i.e., the quality of travel for a user from an origin to a destination.

The "ideal" multimodal system performance measure sought by this research would provide feedback on how travel modes work in concert to serve travelers. It would measure performance from different perspectives, including system productivity, efficiency, and resiliency. It would apply across a variety of settings and locations in the country as well as differing time periods.

The research team quickly found that such a measure would require a new theoretical and analytical framework and new types of data. The new framework would need to orient around a multimodal, complete trip perspective rather than a single mode, facility-based perspective used by most current performance analysis.

Developing and applying complete, trip-based system measures presents several challenges. The first is obtaining complete trip information across all modes and at all times of the day. Some companies are collecting large amounts of complete trip information from cell phone and Global Positioning System (GPS) devices, but do not sell or share the data because of concerns about privacy and incompatibilities with business models. Other companies are expanding smaller samples of complete trip data to simulate travel throughout the day, but those data remain incomplete for a system measure. Despite the hurdles, it is possible to foresee a time in the not too distance future when such data will be readily available.

The second challenge is overcoming institutional and analytical inertia. The transportation profession has a long history and familiarity with single mode, facility-based measurements, due in large part to the availability of facility-based data and to the modal orientation of transportation agencies. Multimodal system performance measurement could fundamentally change both analytical methods and perspectives.

Research Findings

The initial literature review found no "ideal" or universal multimodal transportation system performance measure or approach. It did uncover a high level of interest in pursuing, defining and testing such a method, however.

Multimodal system productivity (MSP) emerged as the "ideal" system measure. MSP is based on the classic definition of productivity: the ratio of inputs to outputs in the production process. For the multimodal transportation system, completed person trips are production outputs and network travel times, or network minutes, are production inputs. The MSP score is the number of completed person trips per network minute. The higher the score, the higher the productivity of the system.

The MSP measure requires completed person trip data, which were not available for the pilot tests. Two surrogate measures, person trips and time-weighted person trips, were used for the three pilot tests in downtown Philadelphia; the San Marco to Escondido corridor north of San Diego; and Crystal City in Arlington, Virginia. The following is a summary of the key findings from the pilot tests.

• The pilot sites were selected because of the availability of multimodal data, yet data gaps from all three would not allow for full measurements. The pilot study data issues reflect the disjointed nature of data collection across travel modes and the need for future initiatives and research on coordinating data collection.
• Feedback from the surrogate measures indicated how different modes performed and hinted at the interplay of travel demand across the modes, but the process confirmed the challenges of using facility-based data and measurements to report on multimodal system performance, primarily due to difficulties with defining the system and aggregating data.
• Without complete trip information and a complete trip perspective, the surrogates did not offer direct feedback from a traveler's perspective.

Innovation Plan

The innovation plan, developed from the findings of this research, includes these proposed initiatives:

• Obtain and improve multimodal system performance measure data – the focus is on either obtaining complete person trip data from sources such as Google or Apple, which have large sample sizes, or improving the expansion techniques for sources with smaller sample sizes. Improvements in travel data, particularly coordinating collection across modes, are also recommended.
• Refine the MSP measure – the types of refinements include further developing the complete trip analytical framework, using complete trip data, once available, to develop and test the MSP, and testing the MSP in a variety of planning, programming, and management and operations applications.

Research Programs and Projects

Recommended research orients around support for innovation initiatives and other related research. For the innovation plan, research programs and projects focus on data development, measure development, and measure applications. Research on the opportunities for understanding relationships with multimodal system performance measures include the areas of policy development, transportation management and operations (focusing on system resiliency), transportation planning and programming, travel demand modeling and forecasting, and relationships with other transportation performance measures.

Summary

The FHWA will consider these proposed innovation and research topics for future funding. As of Fall 2018, FHWA has identified a follow-on study for potential funding.