Simulation of Self-driving System by implementing Digital Twin with GTA5

Computer simulation based on digital twin is an essential process when designing self-driving cars. However, designing a simulation program that is exactly equivalent to real phenomena can be arduous and cost ineffective because many things have to be implemented. In this paper, we propose a method using the online game `GTA5' as a groundwork for autonomous vehicle simulation. As `GTA5' has a variety of well-implemented objects, people, and roads, it can be considered a suitable tool for simulation. By using OpenCV to capture the GTA5 game screen and analyzing images with YOLO and TensorFlow [1] based on Python, we can build quite an accurate object recognition system. This can lead to the writing of algorithms for object avoidance and lane recognition. Once these algorithms have been completed, vehicles in GTA5 can be controlled through codes composed of the basic functions of autonomous driving, such as collision avoidance and lane-departure prevention.

I. Introduction

With the day-by-day development of technologies for deep learning and big-data analysis, applications that use artificial intelligence are rapidly expanding. Among these, the development of algorithms using artificial intelligence is indispensable for autonomous driving. When achieved in real life, self-driving technology will be valuable for reducing traffic accidents, achieving efficiency in time, space and other resources and ease of driving for the disabled. However, as autonomous driving technology has several problems such as stability and safety, virtual-reality simulation is required [2], even essential. In the design of simulation programs, it is difficult to evaluate and implement all occasions and objects in the real world [3]. If one tries to construct a simulation that is close to real phenomena, much money and time is needed to perform the complex manual work. In this paper, we suggest GTA5 based on auto-drive simulation in which the NPC's movements, driving, and road conditions are realistic [4].