Project Title

Multi-city Direct-Demand Models of Peak Period Bicycle and Pedestrian Traffic

Collaborating Universities

Virginia Tech
1424 S Main St.
Blacksburg, VA 24061
540-231-6000
www.vt.edu

Principal Investigator(s)

Steve Hankey (VT) Email: hankey@vt.edu
Ralph Buehler (VT) Email: ralphbu@vt.edu

Funding Source(s) and Amounts Provided (by each agency or organization)

USDOT: $44,616 (Federal), Virginia Tech: $44,616 (Match)

Total Project Costs

$44,616 (Federal)/$44,616 (Match)

Start Date

06/01/16

Completion Date

05/31/17

Description

In order to shift trips to non-motorized modes, transportation planners and engineers need better spatial estimates of walking and cycling traffic as an input to the planning process; for example, to assess exposure to hazards, evaluate infrastructure investments, or locate facilities. Direct-demand models are useful tools for generating spatial estimates of pedestrian and cyclist traffic volumes. Our proposed work aims to address a key issue with single-city direct-demand models of bicycle and pedestrian traffic (i.e., that models are typically not transferable between cities). Our research approach involves mainly three tasks: (1) compile the first multi-city dataset of bicycle and pedestrian traffic counts using the National Bicycle and Pedestrian Documentation Project (NBPDP) count repository, (2) assemble a geocoded, national-level database of predictor variables (e.g., land use, transportation, demographic) at each count location, (3) combine outputs from Tasks 1 & 2 to develop multi-city direct-demand models of bicycle and pedestrian traffic. We expect two novel outcomes from our multi-city models. First, our models will yield more generalizable results about how certain aspects of the built environment are correlated with bicycle and pedestrian traffic (i.e., since our models will be developed from counts in multiple cities across the US). Second, we will be able to generate spatial estimates in communities with few or no counts with greater confidence than previous, single-city models. We expect that our models could be coupled with efforts to tailor non-motorized traffic count campaigns for model-building to develop more robust models in future research.

Implementation

We will explore partnering with FHWA to disseminate our direct-demand models via an existing web-platform called the Non-motorized Toolkit. The Non-motorized Toolkit was originally developed by Jeremy Raw (FHWA) to act as a repository and demonstration platform for bicycle and pedestrian planning tools (e.g., scaling factors, single-city direct-demand models). Our previous models based on Minneapolis count data are currently used on the Non-motorized Toolkit to estimate bicycle and pedestrian traffic volumes in other municipalities. We will explore using the multi-city direct-demand model to replace the Minneapolis model used on the Non-motorized Toolkit.

Impacts

Our goal is to make our direct-demand models publicly available via the Non-motorized Toolkit to enable local planners who have little access to resources for non-motorized travel an easy to use tool for estimating bicycle and pedestrian traffic. These models represent an improvement over existing single-city models.

Web Links to Reports and to the Project website