pacific nw transportation consortium region 10 Pactrans

Using Computer Vision to Evaluate Bicycle and Pedestrian Improvements

PI: Don MacKenzie (UW), dwhm@uw.edu, ORCID: 0000-0002-0344-2344

Co PIs: Mike Lowry (UI)

AMOUNT & MATCH: $180,000 from PacTrans; $180,000 Match

PERFORMANCE PERIOD: 3/16/2022 – 3/15/2023

STATUS: Active

CATEGORIES: Bicycle, Pedestrian, Computer Vision

RESEARCH PROJECT HOT SHEET:

UTC PROJECT DOCUMENTATION:

FINAL PROJECT REPORT: will be available once completed

PROJECT DATA: will be available once completed

DESCRIPTION: The City of Boise, Idaho is about to embark on an ambitious effort to improve safety and mobility for bicyclists with a $180,000 investment along four arterials. The City’s project will convert traditional bike lanes into separated bike lanes by introducing candlestick bollards to physically separate vehicles and bicyclists. This safety improvement is motivated by the City’s recent adoption of Vision Zero goals.

This project will test whether introducing candlestick bollards on bike lanes along four arterials close to grocery stores will increase bicycle trip making for shopping because the bollards will slow auto travel and increase cyclists’ perception of safety. The city selected bike lanes near grocery stores because there is a clear origin-destination pair between neighboring homes and the store and the decision to choose a bicycle doesn’t require a dispersed network to connect the origin and destination.  The team will use computer vision devices to measure the number of bicycles and automobiles, speeds, and separation distance before and after the introduction of the bollards. The computer vision devices will be installed at the improvement locations and at locations that will not be improved for “control group” comparison.

This PacTrans project will develop a repeatable process that cities, state DOTs, and other agencies throughout the Pacific Northwest can use to analyze the large amounts of data generated by modern computer vision devices. The results from this project will demonstrate how the data from this emerging technology can support decision making for other street improvements related to pedestrians and cyclists.