We see on-street parking competition from transit, bicycles, online shopping delivery trucks, shared mobility service companies, and a variety of other usages. People love convenience, but the rigid, daily demand for on-street parking has consequences, including double parking.

Anyone who has driven in a city knows the frustration of encountering a street blocked by double-parked vehicles. Improving enforcement might be one of the solutions to discourage the practice, but knowing where to target is crucial. Researchers at New York University’s C2SMART Center have built a novel data-driven integrated machine learning model for estimating the actual frequency of double parking based on extensive data available in New York City; this random, forest-based, data-driven approach offers an alternative method to estimate street-level double parking activity and identify hotspots.

C2Smart’s researchers break down their research, the resulting data, and what it all means for parking and mobility (including operations that hope to reproduce it all in their areas) in the February issue of Parking & Mobility magazine–and it’s fascinating. Don’t miss it–read the whole story here.