In a recent paper for the National Institute for Transportation and Communities, researchers took a close look at the shockingly nascent science of bicycle counting and how cities could do a better job of measuring where its residents ride…
Unlike previous studies, the data included both physical counters that automatically sense when a person on two wheels is rolling by a specific location, as well as more diffuse data from bikeshare providers, voluntary ride-tracking app Strava, and the Big Data company Streetlight, which uses anonymized cell phone data to automatically sense how many people are pedaling through an entire region in real time…
To get the most accurate possible bicycle traffic forecast, Kothuri and her team developed a “pooled” model that synthesizes data from every available traffic source across the six cities — though she cautions that creating a bespoke, city-specific model using a similar methodology would yield the most accurate predictions, and that local leaders should consult her paper for tips on how to do it right.