Studentenvorträge zu Projekt- und Abschlussarbeiten

 

Alexander Lechner: Erweiterung des REFLECT Media Players zur Darstellung von 2D-Inhalten

Projektarbeit - betreut von Dr. Shadi Al-Dehni, Gilbert Beyer

Die Arbeit beschäftigt sich mit der Erweiterung des Reflect Media Players (RMP) zur Darstellung von 2D-Inhalten. Der RMP ist Teil des Reflect Projektes, das unter anderem an der Ludwig Maximilians Universität München entwickelt wird. Es besteht bereits eine Version des RMP, die in der Lage ist 3D-Inhalte im VRML-Format anzuzeigen und zu manipulieren. Da die Erstellung von 2D-Formaten meist einfacher ist als die von VRML-Dateien, werden im Rahmen dieser Arbeit drei zweidimensionale Grafikformate miteinander verglichen. So soll die Technologie gefunden werden, die am besten dazu geeignet ist in den Reflect Media Player integriert zu werden um auch 2D-Inhalte anzeigen zu können. Nach der Evaluierung wurde in Anlehnung an die bisherigen Anforderungen eine neue 2D-Komponente für den RMP entwickelt.

--------------------------------------

Steffen Ritter: Evaluation of the interactive advertising jingles prototype concerning user interaction

Bachelorarbeit - betreut von Gilbert Beyer

Ziel der Arbeit war es, den Prototyp für Musikinteraktion des Lehrstuhls hinsichtlich seiner Usability zu evaluieren. Dabei wurden zusätzlich zu den vorhandenen Gesten zur Steuerung der Interaktion noch weitere implementiert. Zusätzlich wurde der Anwendung eine geeignete Visualisierung verpasst. Die folgende Evaluation hatte dabei zwei Ziele: das Testen und Bewerten von verschiedenen Gesten für die Musikinteraktion und zu messen, ob sich Nutzer besser an einen Gegenstand erinnern können, wenn die Interaktion nicht nur rein visuell, sondern zusätzlich mit akustischen Elementen erfolgt. Daraufhin wurde eine Usability Evaluation geplant, vorbereitet, mit 21 Personen durchgeführt und anschließen ausgewertet.

---------------------------------------

Marianne Busch: Integration of Security Aspects in Web Engineering

Diplomarbeit - betreut von Dr. Nora Koch

Secure web applications are becoming increasingly important due to rising cybercrime as well as the growing awareness of data privacy. Since adding security features to already existing applications can be a very time-consuming task, it is important to take security aspects into account while planning and modeling a web application. In this thesis, the challenges of security in the web and currently existing security and web engineering approaches are investigated in order to create a comprehensive and coherent security modeling technique for web applications. Web developers should be enabled to model security aspects as authentication, access control and secure connections efficiently. For this purpose, the UML-based Web Engineering (UWE) approach, which was developed at the Institute of Programming and Software Engineering is enhanced. Despite its seamless integration into UWE, the concept presented here can also be utilized independently for other UML-based modeling techniques. In order to create a versatile approach, the functionality of the MagicDraw plugin MagicUWE is complemented by appropriate features. A case study of a Hospital Information System verifies the suitability for daily use of the approach as well as of the tool. To this effect, a prototype is not only modeled, but also implemented. Additionally, an address book example presents alternatives regarding the usage of modeling elements.

 --------------------------------------

Robin Promesberger: Simulation of autonomous robots in a three-dimensional environment

Diplomarbeit - betreut von Dr. Matthias Hölzl

This thesis reports on the architecture, modeling and implementation of a software framework, that enables the user to define, program, and simulate virtual, autonomous robots in a three-dimensional environment. The robotics simulator Robsi presented in this thesis provides an easy to use framework to design, create, and program user-defined simple or complex virtual robots and an interface to simulate them in real time under realistic circumstances in an appealing graphical representation of an environment. It is an open source robot simulation framework in a three-dimensional environment using high-end graphics and physics. Its target audience are researchers and programmers who need a tool to simulate highly customizable and programmable robots to study situated artificial intelligence, swarm robotics, and more generally AI algorithms, in the context of autonomous robotics and autonomous agents. Its intention is to provide a real world simulation, while maintaining a high level of readability and usability. Robsi is a software framework, that enables the user to create his own virtual robots with multiple physical constraints, define and program custom robot controllers, attach sensors, modify the environment, and test the robots in a highly detailed simulation. The user can easily configure, save, and load the parameters of the simulation for various situations to efficiently simulate one or swarms of robots. Based on such simulations, the user can quickly gain a lot of knowledge about the general function of the programmed robots, because the use of highly detailed physics, and the strict separation of the environment and robot programs prove very useful to discover flaws and errors in the basic design and functionality of a robot. The user can then easily iterate the development process, without having to build actual robots. Also, Robsi provides an interface to attach any kind of robot controller to switch from the simulated to actual robots, without having to migrate the robot programs to other platforms. Therefore in actual robot development, the use of the simulator Robsi can save a lot of money and development time.