Workshop Descriptions

Speed Data-ing….for reals!

Back by popular demand is the networking event we call ‘Speed Data-ing for Reals’.  This will be at 5:30 pm on Sunday in the Atrium Ballroom and will provide opportunity for participants to practice interviewing skills, meet new colleagues, chat with existing colleagues and get tips from seasoned veterans on how to network.  It’s a highly interactive session where you’ll move from person to person every few minutes (hence the ‘Speed’ part of the title).  It is a great way to kick off the conference as you will surely come away having met new people to then reconnect with throughout the week.  Don’t Miss This!

VisionEval Tutorial

This tutorial will offer planners and analysts a practical introduction to VisionEval, a strategic modeling framework currently under development through a pooled fund led by seven states and three Metropolitan Planning Organizations.  The moderator and panelists will walk participants through the key elements of choosing to use VisionEval, developing scenarios to analyze, conducting the analysis, and presenting the results.  Participants will learn how to obtain and install the software, and how to access additional documentation, training materials and technical support. Participants will learn how to:

  • Decide if VisionEval is useful for their agency
  • Install the VisionEval system on their own computer
  • Conduct tasks required to develop data inputs and scenarios for analysis with VisionEval (specifically the Regional Strategic Planning Model)
  • Run the scenarios, assemble and interpret the outputs

The tutorial will be based on a real world example of scenario analysis performed by the Atlanta Regional Commission (also documented on the VisionEval website, https://visioneval.org). To illustrate the flexibility of the VisionEval Framework, an additional application context will be described briefly by staff from the Oregon Deparment of Transportation.

The VisionEval tutorial will show you how to download and use VisionEval, and will demonstrate two examples.  Although it is not a requirement to download VisionEval beforehand, the visioneval.org website has a download page.  Here are the steps for a Windows PC.  They should not require administrator access on your computer. 

  1. Make sure you have about 2 GB of free disk drive space on your computer.
  2. R 3.5.2 is a prerequisite for VisionEval.  If you don’t have R 3.5.2 already, download it from here:  https://cran.r-project.org/bin/windows/base/old/3.5.2/.   To install R, run the executable file  R-3.5.2-win.exe
  3. Navigate to https://visioneval.org/category/download.html.   It includes links for downloading R, the VisionEval installer, and some basic instructions for installing and running VisionEval. 
  4. Download the VisionEval installer zip file.  This file is 600 MB.
  5. Unzip the VisionEval zip file to a destination folder of your choice  (needs about  1.3 GB of disk space)
  6. To complete the installation, navigate to that folder, and double-click VisionEval.bat. 
  7. Test that installation has been successful by running a model using the built-in example data, for example, verpat(). See The VisionEval Download page for more details.
  8. If you are already an RStudio user, VisionEval also works in that environment.  Make sure that RStudio is using R Version 3.5.2 (Tools / Global Options / General / R Version).  Once you have run the VisionEval.bat file one time, you can open VisionEval.Rproj to start VisionEval within RStudio.

Tips:

  1. After installation, to start VisionEval, simply run VisionEval.bat again (it loads the needed R packages, and presents you with an R prompt)
    1. If you are using the R GUI, you can turn off the buffering to better see the progress of a model run.  Uncheck Misc / Buffered Output from the top level menu (or press Ctrl-W).

ActivitySim 1.0 Tutorial: An Open-Source Activity-Based Travel Demand Model System

This tutorial will provide an introduction to ActivitySim 1.0, the first release of an advanced, open-source, activity-based travel behavior modeling software based on best software development practices, and available for distribution at no charge to public agencies, and to the public at large.

The ActivitySim project is led by a consortium of Metropolitan Planning Organizations (MPOs) and other transportation planning agencies, which provides technical direction and resources to support project development.  New member agencies are welcome to join the consortium. All member agencies help make decisions about development priorities and benefit from contributions of other agency partners.  The Association of Metropolitan Transportation Organizations (AMPO) facilitates this consortium.

Participants will be introduced to the overall ActivitySim 1.0 model components, inputs, outputs, and sensitivities.  Participants will review data requirements, including network performance data, and will run sensitivity tests on an example data set.  By the end of the tutorial, participants will understand ActivitySim capabilities and requirements for local implementation. 

Urban Freight Delivery Panel

The growth in online shopping has made it possible to bypass traditional storefronts leading to a decrease in shopping trips. However, all this increase in shopping trips has led to an increase in truck traffic in urban areas as good are delivered to residencies. These have an adverse impact on safety, congestion, and pavement lifecycle. Therefore, agencies need to understand the magnitude of this travel in their regions. In this session participants will understand the challenges and what data and changes to the modeling framework are needed to account for urban freight delivery.

Collaborative and Open Source Initiative Governance

Community developed and maintained (i.e. open source) tools are emerging as a best practice for transportation modeling.  Unlike privately owned tools, or tools developed by and for a single user, community-based tools require collaborative governance and stewardship.  This workshop will be a round-table discussion with invited participants representing different collaborative/open source projects to share lessons learned and ideas for improved project governance, followed by a facilitated question and answer session.

The takeaways for the audience include:

  • Knowledge of new easily accessible modeling tools
  • Awareness of partnerships and their value for developing and maintaining agency tools
  • Ideas and suggestions for improving current collaborative projects
  • Understanding the benefits/challenges of Collaborative and Open Source projects

Participating projects will be asked to reference the workshop in their online presence to help keep the discussion going after the conference 

DTA Development and Application Workshop

This workshop will bring together three speakers to present and lead breakout discussions on DTA applications at the operational level, corridor level, and in conducting a regional pricing study; as well as the challenges and solutions to integrating DTA into a regional travel demand model.

Workshop participants will learn the following:

  • How to set up a large model – input data, decisions about zone size (zonal disaggregation), developing trip matrices, time periods – so that it runs in a reasonable time
  • Tips and tricks on model calibration/validation
  • Coding in traffic-related phenomena such as delays due to parallel parking and pedestrian crossings
  • Challenges with modeling large-scale disruptions and pricing scenarios
  • Coordinating the DTA model with signal timing and analysis software (Synchro) as well as regional travel demand modeling software

Participants are encouraged to bring their own DTA experiences, questions, and challenges for discussion during the breakout sessions.

For more information on integrating DTA into a regional model, please see: Forecasting the impossible: The status quo of estimating traffic flows with static traffic assignment and the future of dynamic traffic assignment by Norman L. Marshall https://www.sciencedirect.com/science/article/pii/S2210539517301232?via%3Dihub

Developing and Implementing Sustainable Strategies and Solutions for Multimodal Travel Behavior Data – NextGen NHTS

The workshop has three parts to it, beginning with updates on recent projects related to streamlining the National Household Travel Survey (NHTS) and other state/regional travel surveys to allow for the collection of more frequent survey data and the development of complimentary data products.  Attendees will then hear about relevant projects that illustrate opportunities for advancement before breaking into small groups to identify research priorities and data needs.  In the final portion of the workshop, these priorities and data needs will be discussed, with identification of practical approaches to moving the practice forward. 

This workshop is designed to update attendees on plans to streamline the NHTS design with a combination of more frequent survey data and leveraging passive data products and provide attendees to discuss key design aspects for both FHWA and the general travel survey community.

Transit Visualization Tutorial

We will be examining three different ways of interactive visualization of transit and travel behavior data, with the R Shiny platform. Users familiar with R can install the ‘shiny’ package and begin exploring functionality with the many tutorials here: https://shiny.rstudio.com/tutorial/.

There are public code repositories at https://github.com/metrotransit that we will be exploring during the session. Participants are encouraged to fork these repositories ahead of time to be able to explore the code. They are:

– https://github.com/metrotransit/route-trends to analyze and visualize transit ridership data time series

– https://github.com/metrotransit/speedRT to turn bus location messages into on-street speed visualizations and analyses

Project Forecasting Accuracy Workshop

The National Cooperative Highway Research Program (NCHRP) 08-110 project is addressing a major knowledge gap – the accuracy of traffic projects – through a series of three related podium presentations. A related podium presentation on the forecasting accuracy of transit projects addresses that particular mode. With this background, the NCHRP 08-110 research team recommended that agencies track the forecasting accuracy of transportation projects. There are challenges to doing this, including institutional, technological, and funding-related issues. We propose this session to have an open dialogue with the modeling and forecasting community about (1) the benefits of assessing forecast accuracy and (2) challenges with performing this analysis.

The first part of this session will describe the most recent information on project forecasting accuracy. Presentations will include the findings from the ongoing NCHRP 08-110 study and from personal research from David Schmitt, AICP on the historical accuracy of transit forecasts.

The second part of the session will involve a panel of practitioners and experts in the transportation forecasting field. The panel will discuss major topics such as: how do forecasting customers (i.e., the public and public agencies) view transportation forecasts and forecasters, how can transportation forecasts be communicated better, how to improve the tracking of forecasts, and how to make forecasts better. The panel consists of public agency staff and private sector consultants who regularly deal with and/or produce forecasts for transportation infrastructure forecasts.

CTPP Data and R-based Survey Analysis Workshop

This workshop is in two parts, first a dive into the Census Transportation Planning Products Program and its rich demographic data set.  Pitfalls and caveats of census and census based data, along with a healthy dose of generally good data practices. Then we will look at an exciting way to access the data into R, the opportunities available in a new way to mash CTPP and the National Household Travel Survey (NHTS) along with other regional Household Travel Surveys, including the ability to make customized crosswalk requests from the CTPP and have them publicly available. You will leave this workshop with the R tools to repeat the demonstrated analysis, and directions for incorporating other local regional travel survey datasets.  Intrigued?  You should be!