자율주행자동차의 도입을 위한 논의 -정보교환, 운전행위, 법·제도를 중심으로-

Abstract

In order to commercialize self-driving cars, comprehensive and organic discussions covering various areas are needed rather than a single point of view that they are technical and industrial related to autonomous driving. Along with the development of economy, industry and technology, the relationship of the entire social structure, including legal, administrative and cultural domains, shall be established. In this school, we looked at major issues from a system, social, legal, and institutional perspective for introducing self-driving cars. First of all, the elements needed for autonomous driving will be various information. Information takes place between self-driving cars and various devices, but the exchange of information has security vulnerabilities in itself. In order to address these security issues, it is necessary to ensure safety by continuous updates of the software and to introduce open source software that can improve the quality and security of the software. Second, human driving capacity and driving error were discussed for Level 3 self-driving cars where concepts of support for high-speed driving and self-driving cars that do not presuppose human intervention intersect. Although prior research was insufficient, it emphasized the need for 󰡒standardization󰡓 that had been verified by automobile manufacturers, artificial intelligence, human engineering, brain science experts, and lawyers for 󰡒the discretion of the time of transition and the response of human operation after transition󰡓. Third, the driving algorithm of self-driving needs to be implemented in a way that responds to mutual traffic behavior and recognizes and avoids risks. In addition, it is necessary for autonomous driving systems to lay the groundwork for algorithms that focus on 󰡐Immediate Creativity󰡑 of traffic rules, and if they have such functions, standards for sharing responsibilities among automobile manufacturers, algorithm designers, users of autonomous vehicles, and stakeholders are needed. Therefore, it is necessary to further research from the academic perspectives of computer science, artificial intelligence, humanities and law related to autonomous driving.