Integrated Modelling of Functional Capabilities and Reliability Analysis of Outdoor Autonomous Vehicle Intelligence

Abstract

The autonomous vehicles concept and development were founded in the 1980s, but they became more famous and advanced more than a decade ago. Autonomous vehicles were created due to the advancement of different technologies, and it was believed to portray the progress of the 21st century. This idea led people to think these autonomous vehicles might help reduce or mitigate road accidents. However, firstly, according to the National Law Review, early accidents were recorded, and some were deadly. Secondly, the African continent has been left behind concerning technological advancement; hence, it is currently not ready for so-called smart cities. Therefore, the problem this dissertation looked into is that there is an issue of complexity associated with autonomous vehicles (with independent levels 4 and 5). The study aimed to objectively understudy the reliability of the intelligent autonomous vehicle amidst inter- and intra-complexities associated with autonomous ground vehicle navigation requirements. Therefore, an appropriate methodology had to be selected to fulfill the aim. Thus, two research methodologies were considered for this dissertation, which are (1) design science research and (2) systems thinking methodologies. Additionally, a unification of these two methods was established, and a framework was designed. An optimal physical structure was developed using the established framework and analysing autonomous vehicles’ sensor fusions. Furthermore, the reliability analysis model was formulated. The use of systems and reliability engineering theories and applications were adopted to develop and model the optimal structure and reliability model. Finally, the reliability of the autonomous vehicles with respect to traffic rules was calculated. It was found that there is a 99.94% chance that the autonomous vehicle will fail at least one of the traffic rules in 20 minutes.