An overview of the challenges for developing software within the field of autonomous vehicles

Abstract

Autonomous vehicles are constantly collecting information from their environment, employing reasonably sophisticated cameras and sensors. Applying new software on those vehicles should aim at correcting human errors and gaining new knowledge from all road users. The discussion of autonomous vehicles regarding software from various perspectives, including different knowledge branches such as law, social, economic, design, and ethics. Autonomous vehicles present new challenges with interesting problems in several scientific fields, which are gradually being successfully solved. Social dilemmas, moral and ethical issues show different approaches. There is still an unsolvable problem of decision-making when it comes to the involvement of the ”trolley” problem. Autonomous vehicle software should within the future have flexibility to reconstruct collisions in order that the logic, ethics, and moral algorithm applied may be analyzed. This demands the existence of very good software and hardware technology that can record all the data during a collision. This paper will cover security concerns and current challenges for developing software in the field of autonomous vehicles.