Software transparency as a key requirement for self-driving cars

Abstract

Self-Driving cars is a fast-growing area of study both in academia and industry. It is part of a broader domain which involves the development of software for Highly Automated Vehicles (HAV) and notions extracted from Artificial Intelligence/Autonomous Systems (AI/AS). There are many challenges that must be overcome to deliver self-driving cars in a manner that is readily accepted by consumers and society. Studies have shown that although many people are comfortable with the idea of AI helping them to operate their houses or schedule appointments, not many people are comfortable with the idea of cars being driven by AI algorithms. At the same time, insurance companies are concerned about vehicle liability issues and how to demonstrate who/what caused an accident. We believe that self-driving cars that demonstrate transparency in their operations will increase consumer trust which is pivotal to its acceptance and will pave the way for its commercialization and daily use. In this work, we investigate how to pursue the elicitation and modeling of transparency as a Non-Functional Requirement (NFR) to produce self-driving cars that are more robust.