Proposal for automatic valet parking using learning function

Automatic valet parking using learning function is proposed in this study. The number plate (vehicle registration plate or license plate) of each vehicle is read using a camera when entering and leaving the parking lot. The entering and leaving date and time of each vehicle can be searched by using the plate characters. An image recognition technic and a timer function are used at this time. The license plates are stored in a database. When a customer is entering the parking lot, the reception number and the customer’s vehicle plate number are stored on a magnetic card and handed over it to the customer. When leaving the parking lot, the customer goes to the exit and inserts the magnetic card, and the driver completes the payment process. It is the time when the leaving process is completed. The entering date and time and the leaving date and time are stored together in the database. It is important to shorten the time (leaving time) obtained by subtracting the requested leaving time from the actual leaving date and time in the automated valet parking. The patterns and characteristics of users can be extracted through a data-mining technique by utilizing the created database. Some new values can be added based on the characteristics of the customer’s car using this system. These will enable the functionality and efficiency of the system to be improved. For example, if it is known that a registered customer’s vehicle tends to stay parked there all day after entering the parking lot, the system will park the vehicle in a farthest location from the exit (a location with the lowest priority), allowing other vehicles to leave smoothly. The priority can be given to other vehicles. The vehicles that tend to leave the parking lot in a short time are conversely parked in the highest priority location (closest to the exit). This process can improve the time efficiency of leaving. When the leaving date and time of a vehicle can be known in advance, it can be moved closer to the exit and the time to leave can be shortened. In this study, a simulation system is constructed to improve the efficiency of automated valet parking using the above-mentioned learning function.