PhD students of the Department of Computer Science

"We are building teams of humans and machines." (PhD student Alexander Hagg, Department of Computer Science)

Research in the Department of Computer Science is challenging. In the following, we give a small insight into the variety of topics of the research work of our PhD students and list by whom they are supervised at the H-BRS. Further links lead to research institutes, cooperation partners, publications, etc. (Selection, March 2021).


Iman Awaad
Humans are able to come up with plans to achieve their goals, and to adapt them to changes in their environment, finding fixes, alternatives and taking advantages of opportunities without much deliberation.  Despite decades of research, artificial agents, such as robots, are not as robust or as flexible.  If we look at how we manage to get things done despite the ever-changing environments and our own lack of omniscience we find that this is most often accomplished by making substitutions for missing or unavailable objects and making assumptions about objects for which we have limited information. Enabling service robots, operating in domestic environments to use these two techniques to support human users is the aim of the work of doctoral student Iman Awaad.
Supervisor: Prof. Dr. Paul Plöger (Prof. Dr. Gerhard Kraetzschmar)


Saugata Biswas
Online remote training for assembly, operation and maintenance is advantageous in the industry to save time and money, especially when corporations have locations on several continents (e.g. automotive industry, oil companies, etc.) Currently, remote training experience is limited due to limited camera perspectives and the lack of suitable navigation techniques for the camera view depending on the task. In this PhD project, PhD student Saugata Biswas develops an autonomous camera viewpoint management system using a robotic arm in a multi-camera training scenario. This PhD thesis will focus on improving the online training experience of remote trainees. Since 2019, Doctoral student Saugata Biswas is a scholarshipholder of the Institute for Visual Computing (IVC).
Supervision: Prof. Dr. Ernst Kruijff


Ahmad Drak, TREE
Ahmad Drak
develops a flying robot system that is capable of efficiently exploring the ever-changing environment in which it moves. The result is a wealth of useful information that the system is designed to learn and maximize. Firstly, it shortens the time the robot takes to explore its environment, and secondly, it reduces the energy consumption of the robot system. Since 2018, doctoral student Ahmed Drak is a scholarshipholder of the the Institute of Technology, Resource and Energy-Efficient Engineering (TREE). Further Information.
Supervision: Prof. Dr. Alexander Asteroth


Ruben Gonzalez, ISF
Cryptography, i.e. encryption technology, is used everywhere today. In online banking as well as when opening the car by remote control. However, a new type of computer, the quantum computer, threatens the current generation of encryption systems. In order to remain on the safe side, new cryptographic procedures must be conceived, developed and tested. This branch of research is called Post-Quantum Cryptography. PhD student Ruben Gonzalez is researching how post-quantum cryptography can work on the smallest devices, the constraint embedded devices (e.g. fine dust sensors, credit cards).
Supervision: Prof. Dr. Karl Jonas


Alexander Hagg
Engineers and designers often want to anticipate early on in development processes which potential solutions will meet quality criteria that are important to them, e.g. in architecture, aircraft technology, urban planning or robotics. Because not all criteria can be easily described, such processes are usually divided into different phases. PhD student Alexander Hagg investigates how so-called quality diversity algorithms, which are capable of producing large amounts of good solutions, can be embedded as interactive tools in development processes. This creates an interaction between man and machine, which enables the engineer to discover innovative solutions early in the development process.
Supervision: Prof. Dr. Alexander Asteroth


Carl-Daniel Hailfinger
Data security is an important concern for doctoral student Carl-Daniel Hailfinger. Modern computers and their components process and store data in a way that maximizes the utilization of computing units and throughput. These internal optimization strategies mean that the execution is done in a different processing sequence and that the time behavior in reality can vary in a way that users and programmers do not expect. Although the behaviour (architecture) of e.g. a microprocessor, which can ostensibly be detected from the outside, corresponds to the expectations, a considerable gain in speed is actually achieved internally in the so-called microarchitecture through speculative execution, prediction, intermediate storage and reordering of instructions and data. Attacks such as Spectre, which have become public in recent years, use the resulting side effects to spy on data from another process that is protected against access (side channel / cover channel). Doctoral student Carl-Daniel Hailfinger is researching on the development of novel ways of exploiting such side effects in microarchitecture to spy on or transfer data, on the one hand, and how to protect against such unwanted attacks, on the other.
Supervisor: Prof. Dr. Kerstin Lemke-Rust


Melanie Ludwig
In her research, Melanie Ludwig is investigating how she can model a person's fitness during endurance sports on the computer using heart rate only, i.e. how she can simulate and predict it. Normally, the determination of fitness in endurance sports is associated with complex and strenuous tests, which are particularly difficult to implement in hobby and health sports for many reasons. With her computer models, which are based on everyday sporting activities and heart rate, Melanie Ludwig wants to avoid complex tests and support as many people as possible with individual and health-promoting training. Further Information

Supervision: Prof. Dr. Alexander Asteroth

Alexander Marquardt
Data glasses are glasses in whose field of vision all imaginable information is displayed visually at the same time. However, even modern data glasses only have a very limited field of vision. Unwanted side effects such as distortion, incorrect depth interpretation or poor legibility of information are the result - especially with increasing information density. At H-BRS, common interdisciplinary view management methods are researched and improved. For example, part of the visual digital information is to be converted into audio and vibration stimuli. Alexander Marquardt's focus is on the design, development and technical implementation of this novel, multisensory information supply. His goal is to reduce visual complexity. It also aims to draw the user's attention in the most intuitive way possible to the information that is of particular interest to him or her. Since 2018, Alexander Marquardt is scholarship holder of the Institut for Visual Computing (IVC) in Arbeitsgruppe 3DMi.
Supervisor: Prof. Dr. Ernst Kruijff


Aleksandar Mitrevski
Useful autonomous robot assistants need the ability to modify their behavior if it deviates from what is expected or desired; in other words, they need the ability to deal with execution errors and learn from them accordingly. PhD student Aleksander Mitrevski is developing behavioral models that a domestic robot can acquire through experience with the world and which can be used both to predict errors and to diagnose their causes. In particular, he investigates the balance between modelled and learned behaviour and combines techniques such as learning by demonstration, reinforcement learning, qualitative modelling and logical thinking. The overall goal of his project is to make robots more reliable and thus more useful for practical everyday use.
Supervisor. Prof. Dr. Paul Plöger


Argentina Ortega, Autonmous Systems
In long-term operations, robots repeat their programmed tasks again and again and create new plans each time. One of the research goals of PhD student Argentina Ortega is to investigate how robots can improve their planning by using information from their previous runs. From this information, she creates experience models so that robots can reuse their previous plans in an optimized form. This reduces deployment costs while increasing transparency, explainability and system throughput. Since 2019, Phd student Argentina Ortega is scholarship holder of the equal opportunity center and part of MAS-Team of H-BRS.
Supervisor: Prof. Dr. Erwin Prassler


Christoph Pomrehn
Raman and infrared spectroscopy are two different optical measurement methods. In combination with a microscope, they are used to examine the smallest microscopic samples. The samples are examined with regard to their surface condition, but also with regard to their material composition. Under certain conditions, the resulting images (hyperspectral images) may contain complementary information of the same sample. Christoph Pomrehn develops strategies within the scope of this research topic to obtain this information from the generated images and to make it accessible for computer-aided analysis in selected applications. Since 2018, Christoph Pomrehn is scholarship holder of the Department of Computer Science.
Supervision: Prof. Dr. Rainer Herpers


Sven Schneider
In contrast to industrial robots, we expect household robots to be able to perform gripping movements with their arms depending on the situation. However, they must first learn this ability with the help of models from various disciplines (e.g. mechanics or control engineering). PhD student Sven Schneider makes this interdisciplinary knowledge available to domestic robots by developing application-specific languages. If you like, he is an interpreter and language teacher for household robots.
Supervisor: Prof. Dr. Paul Plöger


Maximilian Schöbel, Autonomous Systems
With the help of example videos, you can teach a computer program how people carry out activities in their home environment. The aim is to "train" robots and enable them to provide assistance in everyday life. Unfortunately, there are not enough such example videos from daily life, because first of all they are considered too boring to be adjusted in large numbers on known video platforms and secondly actions shown in parallel in the video have to be marked by hand before the program can learn them. This is why Maximilian Schöbel's dissertation deals with methods from other fields that require little or no manual labelling. These should then be applied to the recognition of everyday activities in videos. Since 2018, Maximilian Schöbel is scholarship holder of the Graduate Institute und he is part of MAS-Team of H-BRS.
Supervisor: Prof. Dr. Paul Plöger


Sven Seele, Institute for Visual Computing
Virtual environments are artificial computer worlds in which people can learn and train skills for the real world. In such environments, even potentially life-threatening situations can be repeated and varied almost as often as desired without exposing users to real danger. Often other simulated participants (software agents) are also part of the virtual world, supporting or preventing users from achieving training goals. The behaviour of these "agents" can make a significant contribution to making the virtual environment credible and thus enhance the learning effect. Doctoral student Sven Seele is therefore investigating how agent behaviour can be generated by simulating cognitive processes in such a way that the simulation appears as plausible as possible, can be easily controlled and users can experience the simulation interactively at the same time. The modeling of personality, emotions and perception plays a decisive role in this process.
Supervision: Prof. Dr. Rainer Herpers


Katharina Stollenwerk, Informatik, Institute for Visual Computing
Lower back pain is an important issue in modern Western societies. With portable devices, postural changes can be monitored by postural training, the shape of the spine can be measured and the spinal curvature can be reconstructed. PhD student Katharina Stollenwerk focuses her research on the systematic evaluation of (specific) sensor-supported wearables and the reconstruction of spinal curvature as well as the analysis of the recorded data. This improves postural training in back training with reliable and objective measurements of the spinal column shape. Trainers and trainees thus gain a better understanding of their actions or concepts.
Supervisor: Prof. Dr. André Hinkenjann


Santosh Thoduka, Autonomous Systems
Robots are usually programmed for tasks by following a list of actions like moving, looking, picking, etc. When something unexpected happens, a robot often cannot handle the situation because 1) it has not recognized that something has gone wrong and 2) it has not been programmed for the new situation. Recognizing such situations allows robots to decide whether to proceed with the task, inform a human about the problem, or try to solve the problem themselves. Santosh Thoduka's work focuses on using the robot's camera to detect unexpected situations that may occur while the robot is performing a task. Since 2018, PhD studnet Santosh Thoduka is scholarship holder of des Graduate Institute. Further Information. 
Supversivor: Prof. Dr. Paul Plöger


Jan Tolsdorf, Informatik
The ongoing digitalisation and introduction of new information systems in everyday working life means that ever larger amounts of personal data of employees are processed by their employers. This development is particularly problematic with regard to employee data protection when employees have neither sufficient knowledge nor control over the processing and thus the right to informational self-determination as a fundamental element of human dignity is threatened. To compensate for the lack of knowledge and skills in exercising the right to privacy at the workplace, Dokorand Jan Tolsdorf is designing and testing an assistance system in the form of a "privacy dashboard" for everyday work. For its prototypical implementation, design guidelines are to be derived from the mental models and privacy perceptions of employees, with the help of which data processing and data flows in the work environment can be prepared in an understandable way, sensitised to possible infringements of privacy and options for intervention can be shown, so that employees become capable of acting.
Supervisor: Prof. Dr. Luigi Lo Iacono


Christina Trepkowski
Augmented Reality glasses are data glasses in whose field of vision all imaginable information is displayed visually at the same time. This information is intended to improve awareness of certain situations by correct perception, interpretation and assessment of the surroundings. Current AR glasses, however, have one disadvantage: their field of vision is so small that the information they display can obscure critical information from the environment, distract the wearer or overwhelm him with too much information density. Like Alexander Marquardt, Christina Trepkowski is also working on converting part of the visual digital information into audio and vibration stimuli. As a psychologist, her focus is on evaluating, comparing and optimizing these novel methods by developing and applying methods to measure the situational awareness of spectacle wearers. Since 2018, Christina Trepkowski is scholarship holder of the Institute of Visual Computing (IVC) in Arbeitsgruppe 3DMi.
Supervisor: Prof. Dr. Ernst Kruijff


Stephan Wiefling, Informatik
Many of us know the problem with passwords: If they are short, we can easily remember them, but they also make it easier for hackers to access our accounts. Long passwords, on the other hand, are safer, but are difficult to remember. Doctoral student Stephan Wiefling is researching how the security of passwords can be increased without increasing the burden on users. One promising approach is the so-called risk-based authentication (RBA), which is used by large online services such as Google, Facebook and Amazon, but which has hardly been researched despite its great potential. Due to a lack of transparency of the services using RBA, it is hardly used by smaller websites. Stephan Wiefling examines how online services use this technology, how users perceive and use it (usability), and how RBA can be used efficiently in terms of data protection and privacy. The results of the research should provide a full understanding of RBA, which could increase the adoption of the technology and help more websites worldwide protect their users from hackers.
Supervisor: Prof. Dr. Luigi Lo Iacono


Youssef-Mahmoud Youssef
Many industrial applications use distributed robot systems, e.g. in intelligent warehouses, in various areas of logistics or in the disposal of nuclear and hazardous waste. To ensure the reliability of these complex systems, it is necessary to develop intelligent, explainable models that efficiently describe the different behavior of the robots. PhD student Youssef-Mahmoud Youssef is investigating the error detection and diagnosis of distributed robot systems on the basis of explainable hypotheses. Since 2019, Youssef Mahmoud Youssef is scholarship holder of the Department of Computer Science.
Supervisor: Prof. Dr. Martin Müller