A Driver-Side Parking Evaluation Model for Roadside Spaces

As more and more people tend to live in cities, smart parking solutions have been intensively studied. In this article, we identify a crucial disparity between reported parking availability and actual driver arrival times. Traditional metrics, such as recall, while widely used, may not adequately capture real-world scenarios and could negatively impact drivers’ parking experiences. To address this issue, we introduce the driver-side evaluation model (DEM), which integrates a queueing model to more accurately simulate drivers’ waiting times for parking. We validate DEM through three sets of experimental comparisons: evaluating mobile versus fixed sensing methods, examining the impact of varied mobile sensing schedules, and assessing the influence of the queueing model on overall performance. The outcome of these experiments showcases DEM’s effectiveness in evaluating diverse parking strategies and improving behavior prediction.