Here you can find current open projects and theses from our research group. We also supervise theses proposed by students, provided that the topic fits into the portfolio of our research group. If you have an idea for your thesis or would like to have more information, please contact the supervisor in charge. Last Update: September 19th 2023.

Bachelor Theses

RoboCupJunior Soccer Simulation (Lehramt Informatik)

RoboCupJunior ( bietet seit über 20 Jahren die Möglichkeit junge Menschen mit Robotikwettbewerben für Naturwissenschaften und Technik zu begeistern. Die Verein RoboCupJunior Austria und die Technische Universität Graz richten seit 2009 die nationale Meisterschaft aus und unterstützt Schulen, Lehrer*innen und Schüler*innen bei der Teilnahme an den Bewerben.

Bedingt durch die Pandemie konnten in den 2019 und 2020 keine physischen Veranstaltungen durchgeführt werden. Daher wurden viele Aktivitäten auf Online-Events umgestellt. Ebenso wurde der RoboCupJunior online durchgeführt. Die teilnehmenden Schulen konnten entweder über Telekonferenz ihre Roboter vorstellen oder an einem simulierten Wettbewerb für Roboter im Bereich Rescue oder Soccer teilnehmen. Dazu wurde ein eigener Simulator entwickelt, mit dem man Roboter-Fußball-Spiele online simulieren kann (

Ziel dieser Bachelor-Arbeit ist es, ein Konzept zu entwickeln wie der RoboCupJunior Soccer Wettbewerb und die Simulation in Schulen genutzt werden können, um einen zeitgemäßen und interessanten Informatikunterricht zu gestalten. Der Vorteil des Online-Wettbewerbes ist, das die Roboter-Hardware als Hürde wegfällt. Im Zuge der Arbeit sollte auch versucht werden, eine Art Boot-Camp und Testturnier mit Schulen zu organisieren. Diese Arbeit kann auch von einem Team aus 2 Studierenden bearbeitet werden.


Learning Tools and Migration Aid for the new Robot Operating System ROS2

The Robot Operating System (ROS) is an open source set of software libraries and tools for building robot applications. From drivers to state-of-the-art algorithms, and powerful developer tools, ROS provides a framework for the easy integration of robotics software. The ROS software is the main use for the development of robotics applications, a successor framework called ROS2 is the new standard in the robotics community. Due to the wide use of the first version of the framework and the young age of ROS2, published packages are often designed only for the previous version. In addition, fundamental changes have occurred between versions, so that one must learn a new understanding for its use. The goal of this work is to address these two problems. For this purpose first a teaching concept for the new ROS2 Framework is to be provided with which the new functionalities can be obtained. Afterward existing packages of the old version are to be converted exemplarily around the necessary migration steps to examine and identify.  


Seminar/Master Projects

Porting of Autoware Stack

Autoware is an open-source implementation of a complete automated driving stack. It provides a complete tool chain for mapping, localization, and navigation for automated vehicles. It is build up on ROS 2. In this project we are interested to port the Autoware stack to the Mercator robot platform of the AIS research group and to evaluate its performance in automated navigation in urban environments. The main tasks of the project are to become familiar with the Autoware stack and to implement and configure the interfaces to the actual sensor and platform setup of Mercator. If there is interested the project can be extended towards a master's thesis.



Weather Dependent Cost Representation for Routing of a Mobile Robot

Cost-based representations of the environment are frequently used in the path planning domain to obtain an optimized path based on various objectives, such as traversal time or energy consumption. However, obtaining such cost representations still heavily depends on the environment the robot navigates is, particularly in outdoor settings with diverse terrain types and slope angles. Furthermore, depending on the weather condictions, the conditions of the environment might change, and therefore also the capabilities of the robot on certain terrain. For example, while the robot might be able to traverse a grassy area on a sunny day, it might have problems after a heavy rainfall or when there is snow. This topic aims to address this problem by using weather dependent collected data of an outdoor environment to generate a cost-based representation fo the environment which a robot can use to navigate in. As basis, a data-driven approach (Deep Neural Networks) which generates a cost representation for various outdoor terrains on a dry day is provided. Using recorded data on a new weather condition (snow), the cost representation is to be adapted and the knowledge of the traversability of the terrain is to be transferred to the new weather condition.  



Towards Explainable Off-road Navigation Systems

Path planning systems are critical elements for autonomous guided vehicles and provide the foundation for autonomous driving in all kinds of domains. While path planning on roads is usually quite intuitive (e.g., by seeing that the route goes onto the highway and therefore must be the fastest route), routes provided for off-road navigation can be unintuitive and may not contribute to a better understanding of the behavior of the system. Providing explanations and justifications for why a route was retrieved has the potential to address these issues. To provide this, this work should integrate an explanation algorithm based on properties retrieved from the environment and the robotic system. In essence, the explanation algorithm should be able to answer the question "Why is the provided path A the fastest, rather than B, which I expected?" Therefore, retrieved properties about the state (e.g., speed limit of the robot, steepness of the environment, impassable terrain) should be integrated into an explanation. A similar approach, which can be used as a basis for this work, has already been explored by  Alsheeb and Brandao.


Master Theses

Software Engineering Workflow for Autonomous Robots

Autonomous robots are complex systems comprising a number of hardware and software components and showing quite complex behaviors. While in other domains like aerospace or automotive elaborated software frameworks and processes are well established this is hardly realized for the autonomous robot domain. In this thesis we like to investigate what good software development practices exists in other engineering domains and develop a concept for such a software development process to produce better and more reliable software. In particular we like to look into requirement engineering, automate testing of robots and its software, verification and software product lines. This thesis will be conducted in cooperation with Prof. Bernhard Aichernig who is an international expert for formal methods in software engineering.


Reinforcement Learning for Offroad Navigation

Automated navigation in offroad environments such as alpine areas or stone deserts is a challenging task. Usually a layered approach with a stack of a global path planner, a intermediate trajectory planner and a trajectory following controller is used. The task trajectory controller is to steer the vehicle along a basically feasible trajectory trough the environment. Local conditions and difficulties such as medium sized rocks, different surfaces (e.g. sand, gravel), slip, slopes, and uneven ground needs to be considered for successful and safe execution of a trajectory. In this thesis it should be investigated if such a controller can be learned using reinforcement learning. The envisioned use case is the Mecator robot platform that navigates in an alpine environment with various terrains. Starting from simulation a basic controller architecture and learning approach will be developed and trained. Once such a controller is obtained in simulation the results will be transferred to the real robot system in order to safe time and resources in training. The thesis is conducted in cooperation with Eduardo Veas from the Institute of Interactive Systems and Data Science.


Intelligence Test for Autonomous Robots

In classical test theory (aka intelligent test) one aims to estimate the capabilities of individuals on a relative (or absolute) scale using a set if items (tasks) and the rate of solving items by the individuals. The test theory is a well established methodology and uses probabilistic approaches in order to consider the nondeterministic nature of task solving (lucky guesses and careless mistakes). The construction of good and discriminative items with different defined difficulty is a main challenge in this context. While this is well researched in the context of humans there is little work in the context of autonomous robots. We are interested in such items designed for robots because we also want to rate the capabilities of autonomous systems (e.g. robot A is more capable than robot B) but also for predicting  how well a robot will perform in a new environment and task  (new item). In this thesis in cooperation with researchers from psychology we like to develop an approach and a set of test items for tasks of autonomous robots such as navigation or manipulation in order to be able to apply test theory from psychology to robots. In a simulated environment proper items (environments and tasks) will be generated (automatically) and rated according to their difficulty for different robots. This set of validated items can later be used to rate and compare the capabilities of the different robots. This aspect is of increasing interest in the context if the certification pf autonomous systems.


Explanations in Semantic Path Planning

To navigate a robot within its environment, geometric path planners such as A* or Dijkstra are fundamental tools which are widely used. Other planning methods which focus on non-geometric representations such as topological planners or semantic planners are being scientifically researched, but are not yet applicable to general practice. The reason often given for this is the difficulty of representing the planning space, as well as the lack of benefits compared to geometric planners. This work focuses on highlighting and exploiting the advantages of semantic path planners such that they can be used in real-world applications in the near future. By creating a semantic environment and using a semantic path planner, generated paths can be explained in human-like manner to the user in natural language. In case of failures, issues of the environment and/or the path can be communicated to non-expert users helping them to understand the funcionality of the robot. Additionally, using semantic planning, alternative routes and explanations for not taking certain routes can be made transparent. Using the advantage of naturally explaining paths to the user aims to increase trust into the system and to reduce the operators cognitive workload during operation. 



Prof. Dr. Gerald Steinbauer-Wagner