2024 Data Fusion Workshop

This Bicycle and Pedestrian Data Subcommittee in-person workshop was a huge success with over 120 attendees providing input on pedestrian and bicycle data strengths and weaknesses as well as needs for data fusion and the latest research. 

Bicycle and pedestrian counts are foundational for monitoring active travel on a network. However, counts are only available at select locations, leaving volumes on most of the network unknown. Emerging data sources combined with traditional counts may improve network-wide estimation, but questions remain about quality, accuracy, and management and methods to be used for estimating volumes. This workshop will examine data quality and management, explore data fusion, and discuss future needs.

This workshop will discuss bicycle and pedestrian data fusion in three sections as outlined below.  Greg Lindsey (University of Minnesota) will serve as moderator. 

1. Introduction: Shawn Turner, Texas A&M Transportation Institute (TTI), will introduce sources for bicycle and pedestrian volume data including definitions, data types, why we collect data, what and why we want to fuse data, and other foundational topics.

2. Case Studies and Discussion of Data Quality and Management: Phil Lasley, TTI, will start us off with a brief presentation on what goes wrong and what goes right in data quality and management. This will be followed by break out groups by data type. Each group will discuss several questions, including limitations and challenges, steps to improve the data, and how to improve data sharing and integration for that data type. Groups will report back their findings which will be used to compile a list of research needs.

3. Data Fusion Examples and Discussion: The workshop will wrap up with two presentations of current research on data fusion followed by a panel discussion as outlined below.

-          Sirisha Kothuri, Portland State University, will share their work on recently completed project “Exploring Data Fusion Techniques to Estimate Network-Wide Bicycle Volumes” which drew data from Location Based Services (LBS), app, ad bike share, as well as more traditional infrastructure and socio-demographic datasets.

-          Md. Mintu Miah, University of California Berkeley, will share the latest findings from a California project investigating how various datasets (LBS, app, and synthetic data) relate to and may be used with data from permanent pedestrian and bicycle counters.

Panel discussion: Big Data Fusion or Confusion? The presentations will be followed by a panel discussion on the present and future of data integration with panelists from state and federal transportation agencies (Steven Jessberger, FHWA; Josh Roll, Oregon DOT; Brian Wood, Washington State DOT; Peter Ohlms, Virginia DOT), moderated by Greg Lindsey, University of Minnesota. This includes the following panel members: Data vendors will be encouraged to provide their perspectives.

Wrap up: The workshop will conclude with a summary of what we’ve learned and where to go from here by TRB Bicycle and Pedestrian Data Subcommittee Chair, Krista Nordback, University of North Carolina.