Search

Higher levels of autonomous driving are bringing sophisticated requirements and unpredicted challenges. In order to solve these problems, the set of functionalities in modern vehicles is growing in terms of algorithmic complexity and required hardware. The risk of testing implemented solutions in real world is high, expensive and time consuming. This is the reason for virtual automotive simulation tools for testing are heavily acclaimed. Original Equipment Manufacturers (OEMs) use these tools to create closed sense, compute, act loop to have realistic testing scenarios. Production software is tested against simulated sensing data. Based on these inputs a set of actions is produced and simulated which generates consequences that are evaluated. This creates a possibility for OEMs to minimize design errors and optimize costs of the vehicle production before any physical prototypes are produced. This paper presents the development of simple C++/Python perception applications that can be used in driver assistance functionalities. Using ROS as a prototyping platform these applications are validated and tested with "Software-In-the Loop" (SIL) method. CARLA simulator is used as a generator for input data and output commands of the autonomous platform are executed as simulated actions within simulator. Validation is done by connecting Autoware autonomous platform with CARLA simulator in order to test against various scenes in which applications are applicable. Vision based lane detection, which is one of the prototypes, is also tested in a real world scenario to demonstrate the applicability of algorithms developed with simulators to real-time processing

Fahrerassistenzsysteme sind heute zu einer Selbstverständlichkeit im Straßenverkehr geworden. In Zukunft werden diese Systeme das Fahren mehr und mehr automatisieren. Es gibt verschiedene Open-Source-Stacks für automatisiertes Fahren, die wohl bekanntesten sind die beiden der Autoware Foundation.In dieser Arbeit werden die beiden auf dem Robot Operating System (ROS) basierten Stacks der Autoware Foundation beschrieben und getestet. Zunächst wird die LIDAR Perception und Detection von Autoware.Auto in Kombination mit der Umweltsimulation CARLA erklärt und anschließend getestet. Zusätzlich wird der bereits inkludierte Perception Layer um einen V2X-Sensor erweitert und mit der LIDAR Detection verglichen. Anschließend wird der komplette Stack von Autoware.AI erklärt und eine selbstfahrende Demo im Kombination mit CARLA erstellt.Die LIDAR Perception von Autoware.Auto erzielt gute Ergebnisse bei der Objekterkennung, obwohl die erstellten Boundingboxen nicht immer akkurat gesetzt werden. Die V2X Sensor Implementierung empfängt die Position, Orientierung und Abmessungen der Fahrzeuge von der CARLA Simulation. Der Autoware.AI-Stack erzielte gute Ergebnisse in der selbstfahrenden Demo. Es ist möglich, das Fahrzeug mithilfe von Autoware.AI von Punkt A nach Punkt B zu befördern. Zusätzlich verhindert der Stack Kollisionen mit anderen Objekten auf der Straße. Nowadays, advanced driving assistance systems have become a matter of course in road traffic. And in the future, they will automate driving more and more. There are several open-source automated driving stacks available, probably the best known being the two stacks from the Autoware Foundation. In this thesis the two Robot Operating System (ROS) based automated driving stacks of the Autoware Foundation are explained and tested. First, the LIDAR perception and detection layer from Autoware.Auto will be examined and tested in combination with the environment simulation CARLA. Additionally, the included perception layer will be extended by a V2X sensor and compared with the LIDAR detection. And second, the full stack from Autoware.AI is explained and a self-driving demo in combination with CARLA will be set up.The LIDAR perception of Autoware.Auto achieves good results in object detect, although the bounding boxes are not always set perfectly. The V2X sensor implementation receives the position, orientation, and dimensions of the vehicles from the CARLA simulation. The Autoware.AI stack achieved good results in the self-driving demo. The stack navigates and controls the vehicle in CARLA from point A to point B while preventing collisions against obstacles on the road. written by Stefan Teschl Masterarbeit FH JOANNEUM 2021