Document Type


Publication Date



This study’s objective was to use data from existing traffic signal infrastructure to estimate pedestrian volumes. Pedestrian push-button actuations were collected from signal controller logs at 49 intersections in western Oregon and an additional 16 intersections in eastern Oregon. These actuations were then compared to observed pedestrian counts, totaling over 34,000 people, obtained from video recordings. After exploring various options, a simple quadratic relationship was modeled using a single measure of pedestrian signal activity: the number of push-button presses (filtered to remove multiple presses within 15 seconds). The model’s predictions showed a correlation of 0.86 with observed pedestrian volumes and had an average error of ±2.4 pedestrians per hour. These results suggest that existing traffic signal infrastructure data can be used to estimate pedestrian volumes in Oregon with reasonable accuracy. Using such pedestrian volume estimates can lead to improvements in pedestrian traffic monitoring, safety assessments of exposure, and equity and health analyses.


This is the dataset to accompany research to be published in a TBD Journal. Which will be linked upon publication.

A Final Report from TREC will also be published and linked in PDXScholar.

Click on the "Download" button to access the data description and instruction.


This work is marked with CC0 1.0 Universal



SPR 857 (213359 kB)