Examenspresentation av Erik Magnusson och Fredrik Juthe

Examiner: Jonas Fredriksson

Opponents: Philip Johansson and Viet Lee

Abstract:
This thesis investigates the possibility of a centralized electrical/electronic (E/E) and software architecture on a small-scale automotive platform and the possibility to use a digital twin. A decentralized system architecture, used in automotive applications today, is compared to a centralized architecture. The autonomous vehicles of tomorrow will require computing units with a higher computational capacity to be able to process larger amounts of data. Centralized E/E architecture can fulfill these demands, also leading to cheaper development costs, shortening lead times and improving the modularity of software. Furthermore, this simplifies the development of advanced autonomous driving (AD)- and advanced driver assistance-system (ADAS) algorithms that require high computational capacity in real-time.

The functionality of a centralized E/E architecture with a central master computer has been implemented on an autonomous platform successfully according to the requirements. The platform is built on an electric Gokart and is augmented with hardware and software to enable self-driving algorithms to control it. The designed system is modular, scalable and suitable to be used in future research and development. A primitive digital twin to the autonomous platform was proven to be possible using Gazebo and ROS2. This software and a digital twin enable rapid model-based development that can be easily transferred to the physical platform twin. The end result is a small-scale automotive platform currently controlled by a drive-by-wire system. Future research would be to implement ADAS and AD algorithms. Due to the modularity, standardisation and scalability of the system, future implementations of algorithms and hardware can be easily implemented.

Welcome!
Erik, Fredrik och Jonas