Anticipating trends through predictive modeling is essential for effective urban mobility management, enabling data-driven strategies to optimize parking and resource allocation. In this tab, users can explore the outputs of the forecasting model.

On the left-hand side, a selection panel allows users to choose the zone of interest and the week of the year for which predictions are displayed. A mini-map dynamically highlights the elements corresponding to user selections, providing spatial context to the forecasted data. The main content comprises three plots, each displaying week-by-week results:

• Transactions forecast: predicted vs. actual transaction volumes;
• Revenue forecast: predicted vs. actual revenue generated from parking;
• Occupancy forecast: predicted vs. actual street-level parking occupancy rates.

By default, the tool shows general results for the entire city, including real data, the average predicted values, the minimum and maximum time series, highlighting the oscillation range of the data across the city. When a specific zone is selected, the plots update to display localized predictions, allowing users to drill down into results for individual parking meters or streets. These elements are simultaneously highlighted on the mini-map for ease of reference. The comparison between real and predicted data emphasizes the model's robustness and performance, showcasing its reliability in capturing complex urban mobility dynamics.