A Precise Autonomous Parking System Using Ultrasonic Sensors

In this paper, we propose the use of ultrasonic sensors for autonomous parking systems. In general, LiDAR sensors are evaluated to be advantageous in autonomous parking systems, but this work attempted to prove that similar performance can be achieved only with ultrasonic sensors. This work designed a Finite State Machine (FSM)-based control logic to divide the parking process into several stages. The experiment was conducted in a simulation environment with a parking space and a real-world environment, and 1/5 scale vehicle was used as an experimental platform. As a result of the experiment, the system using ultrasonic sensors showed similar performance to that of the LiDAR-based system. Especially in precise parking (parking where vehicles are aligned parallel to both sides of the vehicle, and keeping the distance from both sides uniform), the ultrasonic sensor-based system showed better performance. These results suggest that the ultrasonic sensor-based parking system can be a realistic and cost-effective alternative for the autonomous parking system.