Automated low-cost and real-time truck parking information system : [research summary].

In this project an automated real-time parking information system was developed to improve truck-parking safety through the efficient gathering and disseminating of information regarding the use of existing parking capacity. The system consists of four main components: sensing, data collection, data processing and user interface (UI). A pilot deployment was conducted on a Maryland State Highway Administration truck parking facility located on Interstate-95 northbound prior to Maryland Route 32. During the testing period of January 6 through 14, 2013, 1,239 events were detected by the system. Each event refers to any truck arrival or departure activity in a spot. The average overall detection error was 3.75 percent and the maximum error was 5 percent. The error rate can potentially be reduced by using more sensors at each spot and using repeaters to avoid signal blockages. Unlike imagery based methods, magnetic truck detection is completely anonymous and thus privacy of drivers is not compromised. It is also independent of parking layout. In addition to providing real-time parking availability information to the truckers, analysis of historical data for each spot and for the parking facility as a whole can reveal dynamics of events and facilitate efficient operations.