[Back]

M. Karan:
"Using Mobile Edge Computing Technologies for Real-Time Cornering Assistance";
Supervisor: H. Truong; Institute of Information Systems Engineering, Distributed Systems Group, 2018; final examination: 2018-02-26.

Today, modern cars with advanced driver assistance systems (ADAS) and even fully autonomous driving software are on the rise and expectedly will reduce accidents in traffic in the future. But as of now, the vast majority of drivers still drive cars without these technologies. Until cars will be driving fully autonomous, human errors are going to lead to thousands of car accidents. Statistics show that one major cause of accidents is related to cornering. To contribute to safer driving in any type of car, in this thesis we introduce a novel system that assists drivers in real-time while cornering. The system is designed in a way that it can be deployed to both the cloud and/or the emerging edge-computing infrastructure. The goal of edge-computing is to move computational units closer to where the data is produced. We also contribute solutions to processing location data, using and combining external data sources and calculating curves and properties. As a result, we design and implement a prototype that is tested against simulated traffic scenarios. In the evaluation we show how the prototype behaves in terms of performance and data quality. Based on the prototype, we finally demonstrate how much it would approximately cost to serve all drivers across Austria.