Economics of smart parking: predictive analytics in optimizing urban parking revenue

During this era of rapid urbanization and technological strides, parking management has evolved into a crucial facet of urban planning. Smart Parking Systems (SPS) are at the forefront of enhancing parking convenience and reshaping revenue streams for sustainable city living, yet despite their acknowledged advantages, there remains a notable research gap in quantitatively understanding their impact on parking revenue generation. This study addresses this gap by employing advanced predictive analytics to forecast parking revenues and assess the influence of SPS on financial outcomes. The methodology was tested in an urban university campus context, specifically focusing on the University of Texas at Arlington. This university, located in the heart of Arlington, Texas, presents a dynamic urban environment with significant daily traffic flow and parking demand. The model was carefully refined through extensive hyperparameter optimization, cross-validation, and thorough performance testing on both the training and validation datasets and was then evaluated, using data collected after the implementation of the SPS, to assess its impact on parking revenue. The model showcased high forecasting accuracy, as evidenced by a robust coefficient of determination for both training and validation datasets, and the evaluation revealed its strong predictive capability, with actual revenues frequently surpassing predicted ones, underscoring the efficacy of SPS in enhancing revenue. The findings of this study will equip parking lot managers and researchers with the confidence to implement related systems that will enhance the efficiency of their operations and increase their revenues.