Dashboard Cameras Combined with AI Provide an Affordable Method for Identifying Curb Usage

The increasing reliance on transportation network companies (TNCs) and delivery services has transformed the use of curb space. The curb space is also an important interface for bikeways, bus lanes, street vendors, and paratransit stops for passengers with disabilities. These various demands are contributing to a lack of parking, resulting in illegal and double-parking and excessive cruising for spaces and causing traffic disturbance, congestion, and hazardous situations. How cities manage this public asset to support safety and the local economy relies on first understanding the usage patterns over time and space. Various curb management models, such as dynamic parking pricing, and parking reservations have been developed to improve curb usage, but implementing an adequate control policy requires extensive parking analysis to identify area-specific curb characteristics, spatiotemporal demand distributions, common reasons for congestion, and other parameters. Currently, no systematic process for identifying these patterns exists, but emerging machine learning technologies combined with low-tech dashboard cameras mounted on vehicles that routinely travel the area provide a potential, affordable, and scalable method to monitor usage.