Enhancing Autonomous Driving: Evaluations Of AI And ML Algorithms

Abstract

However, the development of fully autonomous driving, as anticipated shortly, still poses several partly unsolved computational, technological, and physical problems. The main one is how to provide vehicles with cognitive abilities capable of understanding the environment. In this work, we consider cloud computing-based solutions that enable autonomous driving tasks to offload the data processing and model predictions from on-board cameras to be processed at remote data centers. Despite several advantages inherent in this approach, the possibility of applying it in practice has not yet been proven, and the discussion of the approach in the context of autonomous vehicles is still lacking in the literature. Consequently, the main aim of this work is to provide independent research on the benefits and feasibility of adopting edge computing data processing in autonomous vehicles, with a focus on its practical implications for the automotive industry and computer science, thereby paving the way for future advancements in these fields.