Predicting Available Parking Slots on Critical and Regular Services by Exploiting a Range of Open Data

Looking for available parking slots has become a serious issue in contemporary urban mobility. The selection of suitable car parks could be influenced by multiple factors-e.g., the walking distance to destination, driving and waiting time, parking prices, availability, and accessibility-while the availability of unused parking slots might depend on parking location, events in the area, traffic flow, and weather conditions. This paper presents a set of metrics and techniques to predict the number of available parking slots in city garages with gates. With this aim, we have considered three different predictive techniques, while comparing different approaches. The comparison has been performed according to the data collected in a dozen of garages in the area of Florence by using Sii-Mobility National Research Project and Km4City infrastructure. The resulting solution has demonstrated that a Bayesian regularized neural network exploiting historical data, weather condition, and traffic flow data can offer a robust approach for the implementation of reliable and fast predictions of available slots in terms of flexibility and robustness to critical cases. The solution adopted in a Smart City Apps in the Florence area for sustainable mobility has been welcomed with broad appreciation or has been praised as successful.