Designing a Requirements Elicitation Approach for Intelligent and Interactive Systems in Autonomous Vehicles

Abstract

Software is becoming an increasingly large part in the automotive industry. From vehicles being connected to each other through networks to artificially intelligent algorithms that aid people while driving, software is quickly becoming an essential part of the vehicle and it will only keep growing as we steer into the age of autonomous driving. Because of this, a thorough requirements specification is needed in order to understand what drivers need from the intelligent systems in their vehicles, but also what system needs have to be fulfilled for the vehicles to reach its full potential. There is not enough research about how requirements can be elicited in the context of intelligent and interactive systems in autonomous vehicles. In addition, there is no proof that traditional requirements elicitation techniques work for these types of systems. The aim of this thesis is to investigate the issues and challenges of traditional requirements elicitation techniques in the context of intelligent and interactive systems in autonomous vehicles and based on the results, propose an approach for eliciting requirements. In order to achieve this, we performed three iterations using the Design Science Research (DSR) methodology. In each iteration, we started with discovering the problems that we are facing right now. Then, we produced an artifact that would address those issues and finally, we would evaluate our artifact. Using this approach helped us propose a requirements elicitation approach consisting of a process model for innovative concepts and a requirements elicitation tool.