Challenges and Opportunities in Smart Parking Sensor Technologies

Abstract

Parking a vehicle in densely populated traffic environments often results in prolonged driving time as drivers search for available spaces, contributing to congestion and environmental pollution. Inefficient parking behavior can be attributed, in part, to the lack of guidance information regarding vacant parking spaces. The implementation of smart parking sensors and technologies offers a solution by guiding drivers to free parking spaces, thereby enhancing parking efficiency. However, it is noteworthy that there are currently no such sensors or technologies employed for open parking lots. This article undertakes a comprehensive review of the existing literature on the utilization of smart parking sensors, technologies, and applications while assessing their suitability for open parking lots. Magnetometers, ultrasonic sensors, and machine vision represent some of the commonly employed sensors and technologies in closed parking lots. Nevertheless, this article recommends the utilization of a combination of machine vision, convolutional neural networks, or multi-agent systems for open parking lots due to their cost-effectiveness and resilience to diverse environmental conditions. While a few smart parking applications indicate the locations of typical open parking lots, none of them provide real-time parking occupancy information, which is crucial for guiding drivers along the shortest route to available space. Consequently, further research in the domains of deep learning and multi-agent systems is essential to develop smart parking applications specifically tailored for open parking lots.