Offene Themen für Projekt- und Diplomarbeiten

Automotive Infotainmentsysteme nehmen immer mehr Funktionalitäten an. Das Volumen und die Komplexität ihrer HMIs (Human-Machine Interface) haben s ich in den letzten Jahren signifikant erhöht. HMI-Fehler sind ein wesentlicher Bestandteil aller Infotainmentsystem-Fehler. Aufgrund der hohen Zahl an Ha rdware, Software-, Länder- und Sprachen-Varianten eines Infotainmentsystems ist eine hohe Testabdeckung das gemeinsame Ziel vieler Automobil-Hersteller. Modell-basiertes Testen stellt eine gute Lösung dar, um die Testabdeckung zu erhöhen und gleichzeitig die Kosten akzeptabel zu halten.

Es entstehen im Rahmen einer Doktorarbeit bei AUDI AG Themen, die für ein Praktikum, eine Projektarbeit oder eine Abschlussarbeit geeignet sind. D ie Verbindung der Themen ist für ein Praktikum oder Projektarbeit und anschließenden Abschlussarbeit besonders geeignet.

Aufgabenbereiche:

• Design von HMI Spezifikation- oder Entwicklungsmodellen
• Modell-basiertes Testen für Infotainmentsystem-HMI

Gewünschte Vorkenntnisse:

• Allgemeine Kenntnisse über Modell-basierte Entwicklung bzw. Modell-basiertes Testen
• Allgemeine Kenntnisse über HMI bzw. GUI Design
• Programmierung mit Java
• Optional: Graphentheorie

Durchführung:

• Vollzeit
• möglichst bei AUDI in Ingolstadt

Introduction

ShoeSense is a wearable interface for mobile devices that consists of a shoe-mounted depth sensor pointing upward at the wearer. Shoesense can recognize a range of hand gestures, that enable the user to quickly perform frequent operations without reaching for the phone, discretely perform operations without disturbing others and enhance operations on mobile devices while supporting accessbility.

Research Question

In this Projektarbeit / Bachelor’s Thesis you will have the opportunity to implement a series of demonstrator applications for the existing ShoeSense prototype that is due to be presented at the ACM SIGCHI Conference on Human Factors in Computing Systems (CHI) 2012 in Austin, Texas. To accomplish this you are also required to extend and refine the existing prototype to be presentable at that venue.

The demo applications are intended to further underline the feasibility of our approach and may also be used to assist field studies of our system while on demonstration at CHI.

Tasks

While working on this thesis, the student will be expected to complete the following tasks:

• Think of creative scenarios for the ShoeSense prototype. This phase can consist of conducting user interviews, scenario writing and storyboarding as well as the development of paper prototypes for your applications.
• Review the physical construction of the ShoeSense prototype and improve upon it if necessary.
• Implement the 2–3 best ideas as demo applications for ShoeSense.
• Think of ways the demonstration application can be incorporated into an on-site field study at CHI.

Requirements

• Student in Media Informatics, Computer Science, Electrical Engineering or related field
• Programming Skills in an object-oriented programming language (i.e. C++, C#, Java, Python)
• Student in Media Informatics, Computer Science, Electrical Engineering or related field
• Experience in computer vision (OpenCV, Point Cloud Library) or embedded systems will be useful but not mandatory.

Basic Planning

The thesis / projektarbeit should be completed within a three month (approx. 12 weeks) time frame:

• Literature review (1 week)
• Familiarization with existing code base (2 weeks)
• Scenario and paper-prototyping (2 weeks)
• Improving hardware prototype (2 weeks)
• Coding of demo applications (4 weeks)
• Writing of documentation and final report (2 weeks)

Visual similarity is the most important factor people use to judge the authenticity of websites. Fraudulent websites like Phishing websites therefore try to closely mimic an original website when attempting to steal data from the respective users. Other means of automatic protection on a non-visual layer often fail because attackers are able to modify things like HTML-Code with the appearance of a website staying the same.

With a given detector framework for visual similarity between websites it is still not clear how to guide the user whenever a possible attack is detected. Similarity between websites ist just measured as probability score. In the end the user has to decide whether an attack found is a false positive or not.

Goal of this work is find a possible user interface for displaying website similiarties and potential attacks to the user. The most promising drafts shall be implemented as a prototype that will be validated during a field study.

• Find related work to the topic of web security UIs and website similarity
• Create UI darfts and conduct a focus group to refine those drafts
• Build a Firefox plugin that is able to use a similiarty backend server to create warnings when necesarry
• Create a framework for logging plugin events and decisions
• Prepare and conduct a field study using the plugin and the data collection server
• Analysis of the results
• Written thesis and presentation of the work