VisionEval A programming framework for disaggregate strategic planning models

Strategic planning is becoming increasingly important as a means to help state and metropolitan area governments select policies and actions to address pressing issues that are fraught with uncertainty. More specifically, Federal direction has challenged state, regional, and local transportation agencies with measuring the outcomes of decisions through performance-based planning, including considering how transportation solutions may impact future goals such as sustainability, health, mobility, etc. Further complicating matters, plans need to be resilient to changing transportation and land use trends and the implications of emerging technologies and constraints. VisionEval is an open source common framework building on the successful GreenSTEP family of strategic planning tools that is intended to address these needs.

VisionEval is a recently established partnership merging this family of tools into an open-source project with a supporting community forum of partner agencies and others sharing in its use and enhancement. The goal is to support a broad array of potential tool uses and enable pooled enhancements expanding the types of outcomes measured or refine the specificity of transportation and land use solutions considered. The work to date by the founding FHWA-OregonDOT partnership has focused on the technical components of putting the various tools on a common programming framework. Bolstered by interest at an RPAT Peer Exchange in October 2015, efforts are underway to build a supportive community around the tool, drawing from successful past and interested future users nationally, who will both define the policy needs and uses of these tools, and set their direction moving forward.

NEW Join VisionEval Project’s next chapter! The FHWA-hosted pooled fund project, Collaborative Development of New Strategic Planning Models, is now open and accepting contributions: Pooled Fund Project Website

Explore the information below to learn more about VisionEval and contribute to our ongoing efforts.

The family of VisionEval tools began with a tool designed to measure GHG emission outcomes in Oregon (GreenSTEP). It was rebranded the Regional Strategic Planning Model (RSPM) when broader measures such as health, household costs, and other outcomes were added. The tool was adapted nationally in modified form as RPAT (Rapid Policy Assessment Tool) and EERPAT (Energy and Emissions Reduction Policy Analysis Tool). This tool evaluates transportation investment and policy strategies such as increased transit service, pricing, parking management, etc. combined with land use plans to assess how different combinations of investment and policy actions impact travel behavior and budgets and thus influence desired community outcomes/goals. The short runtimes enable thousands of scenarios to be evaluated testing plan resilience and has resulted in several web-based interactive visualization tools to explore policy/investment trade-offs with stakeholders and the public.

Oregon has found significant value in RSPM to support performance-based planning in developing future plans, and most MPOs in the state are using the it (and in some cases RPAT), with future interest in enhancements. Several current and past pilot users of RPAT and EERPAT across the country continue to find value from the tool in greenhouse gas planning efforts. This project builds on these collective achievements and allows pooling of resources and energy that will result in better tools, increased user base, a larger pool of technical developers, and a community forum sharing best practices in how the tools are supporting land use and transportation decision-making conversations across the nation.

The core of VisionEval is a common programming framework that allow the four strategic planning models in the GreenSTEP family to share a commmon platform, facilitating development and deployment to a broader community. The framework enables statewide, regional, or urban area models to be built by connecting together modules that are distributed in standard R language packages. The framework establishes interface standards for modules and manages the data flow between modules and a common datastore. Although the inspiration for the VisionEval framework is a set of related transportation and land use models (see below), the framework can be used for other modeling and research purposes as well. The VisionEval project repository on GitHub is the source for project documentation and code. A description of the impetus and vision for the framework is included in the repository here. Several partners and technical contributors have participated in the development of the VisionEval framework to date. Partners include the Federal Highway Administration. Technical contributors include Oregon Systems Analytics (OSA), Resource Systems Group (RSG), Portland State University (PSU), University of British Columbia (UBC), Lane Council of Governments (LCOG), and Volpe.
Strategic planning is becoming increasingly important as a means to help state and metropolitan area governments select policies and actions to address pressing issues that are fraught with uncertainty. Past transportation and land use trends can no longer be assumed as givens and the implications of emerging technologies need to be considered. Furthermore, substantial environmental constraints need to be addressed. NCHRP report 750 describes in detail a number of strategic planning issues that are facing transportation. The objectives of strategic planning are to consider the prospects of future trends, evaluate the potential for influencing those trends and their effects, and develop a strategic vision for the future. A strategic vision identifies desired performance goals and general strategies for achieving those goals. Strategic planning is also sometimes referred to as scenario planning.
Strategic planning models evaluate the consequences of future scenarios that vary with respect to external trends and proposed policy responses to those trends. Strategic planning models are designed to address a wide range of trends and policies, rather than focusing on details. They are designed in this way so that many alternative futures and policies can be evaluated, enabling planners to reason more effectively about intended and unintended consequences of prospective policies. Disaggregate strategic planning models work by simulating the effects of trends and policies on individual households. This has been found to be an efficient way to model the interactions between policies and their potential effects on households having different characteristics and located in different areas.
The impetus for developing the VisionEval framework was to make the code for several related disaggregate strategic planning models easier to maintain, modify, extend, and interchange. Following are brief descriptions of these models. Click on the model name for more information about the model.
The Greenhouse gas Strategic Transportation Energy Planning model was the first of these models to be developed. It was developed by the Oregon Department of Transportation (ODOT) to assist in the development of plans to reduce greenhouse gas (GHG) emissions from light-duty vehicles to meet statutory goals. GreenSTEP models the effects of many different factors (e.g. transportation supply, prices, land use, etc.) on household vehicle ownership and use, and the effects on emissions, traffic congestion, etc. Technical documentation for the GreenSTEP model is available here.
The Regional Strategic Planning Model was developed by ODOT as an offshoot of the GreenSTEP model to support the preparation of metropolitan area strategic assessments and scenario plans. It is a revision of an earlier model named "Metropolitan GreenSTEP". The name was changed because the model (like GreenSTEP) does much more than estimate GHG emissions. A GitHub repository contains the version 3.5 code and sample files. It also contains technical documentation for the RSPM and GreenSTEP models.
The Energy and Emissions Reduction Policy Analysis Tool was developed to "assist state transportation agencies with analyzing greenhouse gas reduction scenarios and alternatives for use in the transportation planning process, the development of state climate action plans, scenario planning exercises, and to measure the reduction potential of various transportation strategies to meet state greenhouse gas reduction goals and targets." The EERPAT was developed from the GreenSTEP model and was revised to have a graphical user interface and to serve all 50 states.
The Rapid Policy Analysis Tool was developed under the 2nd Strategic Highway Research Program that was authorized in 2005 as part of the federal SAFETEA-LU legislation. The model was developed to help planners evaluate the potential effect of smart growth policies on regional travel. Portions of the GreenSTEP model were used in RPAT, but substantial revisions were made to the code. The development site for the RPAT code is here. The RPAT describes land uses with place type categories. A description of place types and ODOT's classification method is available here. ODOT has also developed a viewer for visualizing the place type classifications for an area, an example of which can be viewed here.
A benefit of using strategic planning models is that they enable a large number of scenarios to be evaluated. This provides planners with more information about the potential consequences of future trends and the policies that might be put in place to influence those trends and their effects. This benefit also poses a challenge in regards to how to make sense of the large amounts of output data that are produced. Fortunately, the worldwide web and programming tools developed for programming web content assist with meeting the challenge and make it easier than ever to make model results available to planners and the public. Following are examples of model results visualizations that have been developed.
ODOT Scenario Viewer
ODOT's scenario viewer is a web application that enables users to explore the results of modeling hundreds of scenarios in an interactive manner. Users can investigate how changing future assumptions affects outcome measures. They can also work backwards and determine what inputs are associated with outcomes they choose.
Future Builder
Future Builder is a web application developed for the Central Lane Scenario Planning effort. It enables the public to choose various types and levels of actions for reducing greenhouse gas emissions from light-duty vehicles, and see how those choices affect various performance measures. It also enables users to provide feedback to planners.
Choices and Voices
The Delaware Valley Regional Planning Commission created a web application that allows the public to provide feedback on how the greater Philadelphia region should grow and how the transportation system for the region should be funded, managed, and improved. It provides feedback on how those choices affect various performance measures and allow users to submit their choices to planners.
The rationale and design for the VisionEval framework are documented by papers included in the project's GitHub repository they include the following documents. Check back frequently for the latest documents because the framework is still under development.
VisionEval Strategic Tools for Performance-Based Planning
This paper provides an overview of VisionEval as a model system for implementing analysis tools to support long range performance-based planning.
VisionEval: A New Framework for the GreenSTEP Family of Models: Overview and Approach
This paper describes the rationale and project approach.
VisionEval Model System and Framework Design
This paper goes into more depth on how the framework fits into a model system and the design of that system and the framework.
TMIP Webinar: Introduction to GreenSTEP, the RSPM and Related Strategic Planning Models.
This is the first of a three-part webinar series on the GreenSTEP and RSPM models. This webinar focused on:
  • The motivation for building the GreenSTEP model and the policy issues it was built to analyze.
  • Design overview and major features.
  • Other related models (EERPAT, RPAT) and their differences.
  • GreenSTEP and RSPM application examples.
  • How the GreenSTEP and RSPM models are run.
  • How to obtain the code and technical documentation for the GreenSTEP and RSPM models.
TMIP Webinar: Design and Estimation of the GreenSTEP and RSPM Models.
This is the second of a three-part webinar series on the GreenSTEP and RSPM models. This webinar focused on:
  • Integrated model development process using R.
  • Data and data preparation used in model development.
  • Development of the household DVMT model.
  • Development of the light-weight vehicles model.
  • Development of the PHEV model.
  • Development of the household budget model.
  • Development of the congestion model.
TMIP Webinar: VisionEval - A Modeling System for Building Strategic Planning Models.
This is the third of a three-part webinar series on the GreenSTEP and RSPM models. This webinar focused on:
  • The reasons for creating VisionEval.
  • Model system objectives.
  • Overview of the model system architecture.
  • Key principles that enable 'plug-and-play' operation.
  • Module structure and the packaging of modules.
  • What happens behind the scenes when a model runs.
  • How to download and use the code.
  • Next steps.
  • FHWA: Brian Gardner, Eric Pihl, Jeremy Raw
  • Oregon DOT: Tara Weidner, Amanda Pietz, Brian Dunn