Automatic car parking system

Abstract

The purpose of this project is to leverage on the vast potential of 5G technology for accurate identification/detection a vehicle’s whereabouts and also its parking activities. These parking activities includes both parallel and vertical parkings. The project is formulated in way that the team focuses on enhancing the accuracy of positioning for auto-charging of vehicles at URA's kerbside car parks using the 5G system. In a previous attempt with Differential GPS, accurate positioning was achieved for parked motor vehicles, but failed in an urban canyon environment. Therefore, the current trial uses the 5G system to investigate if it is able to provide better positioning accuracy, with the help of machine learning algorithms. The algorithm's remarkable abilities were shown in the study's findings, as it was able to detect the location of vehicles with a splendid accuracy of 95% within a one-minute timeframe. To confirm the accuracy of the detection, ground truth data from the Global Positioning System (GPS) was utilized. These promising results obtained from this trial will be analyzed to determine if this approach can be implemented in an urban canyon scenario for auto-charging of vehicles and precise positioning for autonomous vehicles or unmanned aerial vehicle (UAV) for logistics delivery. The success of this trial will pave the way for further evaluations of the feasibility and implementation of this technique.