Smart and Cooperative Truck Parking Information Management System

Abstract: The trucking industry is an essential component of freight transportation in the United States, yet it faces significant challenges related to truck parking availability. The insufficient parking spaces, combined with the absence of real-time parking availability information, introduces uncertainty into trip planning and can lead to illegal parking or driving beyond legally allowed hours, both of which raise safety concerns. In 2021, the Washington State Department of Transportation (WSDOT) initiated a pilot project to develop a comprehensive truck parking information management system (TPIMS). This system combines a parking detection technology that collects and analyzes real-time parking data with a predictive algorithm that forecasts future parking availability. From the pilot study, a space-by-space occupancy detection system was chosen as the primary detection. While this system is typically accurate and reliable, it comes with high costs and a complex installation process, as sensors must be installed underground at each parking space. For statewide implementation, however, an entry/exit sensing approach has been adopted at select locations due to its cost-effectiveness and suitability for various parking lot configurations. While this method reduces installation costs, it is prone to error accumulation from miscounts, requiring manual corrections and timely calibration to prevent growing inaccuracies. To address these issues, this study proposes a framework for a human-machine cooperative monitoring and calibration algorithm, enhanced by machine learning-based sequence classification. The system includes an anomaly detection algorithm that regularly compares the entry/exit counting data with ground-truth data. Additionally, a periodic calibration framework, powered by machine learning, continuously monitors the system with ground-truth data extraction components, automatically identifying and correcting common errors. This approach not only mitigates error accumulation and ensures consistent system performance but also minimizes the need for manual intervention, thereby enhancing cost-efficiency. With high-quality data, the statewide deployment of TPIMS will expand access to accurate parking information for drivers, enabling more informed parking decisions. This will help reduce unauthorized truck parking, decrease the number of trucks searching for spaces, and lower the risk of drivers exceeding Hours-of-Service limits, ultimately reducing the likelihood of fatigue-related accidents.