Fit4Ukraine

As a master's student in the double degree program at Anhalt University of Applied Sciences in Köthen, I recently completed my master's thesis entitled "Drift Diffusion Model in Augmented Reality". This project was an incredible opportunity to explore how augmented reality affects human cognition and improves visual perception and reaction times in fast forced decision tasks.

Designing and developing the AR application for both generations of Microsoft HoloLens was a particularly unique experience. Using the Unity Engine, I implemented marker-based tracking, created the required graphics and built a client-server architecture to ensure smooth communication. This allowed me to improve my technical skills and demonstrate the potential of augmented reality to create engaging, innovative and interactive experiences.

The experimental part of the study conducted was equally exciting. With the support of my supervisors Prof. Dr. Johannes Tümler and Prof. Dr. Stefan Schlechtweg, four different scenarios were designed, for each of which 15 participants were recruited. The study provided valuable insights into the effects of augmented reality on cognitive processes, highlighted the specifics of augmented reality in enhancing user experiences in different applications and offered promising directions for future research and development in this area.

As I reflect on my journey, I am grateful for the access to cutting-edge AR hardware and the supportive environment at the xR Lab. The encouragement and guidance from my supervisors were crucial to navigating the complexities of my research. This experience not only sharpened my technical skills, but also helped improve my organizational, presentation, and project coordination skills.

Topic: Investigation of Control and Data Transfer Algorithms between Augmented Reality and Robotic Systems

I built and tested an AR interface that lets a user steer a Double 3 telepresence robot using a smartphone. The AR app (Unity + ARCore on Android) supports touch input and spatial marker recognition (QR codes). Under the hood, I wrote a two-way socket protocol that talks to Double 3’s SDK (the robot exposes a d3 system service over a Unix domain socket), so commands and telemetry flow in real time. 

A key challenge was getting two independent devices to agree on where they are. I solved that with a mutual localization algorithm: the robot acts as the fixed anchor; the phone finds a QR marker on the robot, aligns its AR world to the robot’s frame, and then keeps both frames overlapped by continuously updating the virtual marker from robot odometry. That gives the app reliable waypoints the robot can actually hit.

To evaluate the system, I ran two scenarios on a test site with repeated runs: (1) baseline motion along a planned trajectory without AR, and (2) the same trajectory driven via the AR interface. I measured accuracy (deviation from target) and repeatability across 20 trials per target.

Why is this necessary?

Offices and warehouses: quickly show the way, conduct a tour or inspection remotely.

Education and museums: a guide or teacher can conduct a ‘live’ session from anywhere in the world.

Services and medicine: remote rounds, tips, consultations — when it is important to be ‘on site’ while remaining at a distance.

What the tests showed

The interface is intuitive and user-friendly — the entry threshold is low, and anyone can operate it. To be honest, AR control is still inferior to classic control (when the robot moves strictly according to internal landmarks) in terms of accuracy. This is to be expected — and it is clear where improvements can be made.

Why this still matters: the interface worked end-to-end: anchor with QR, place waypoints in AR, stream them over sockets, and drive the robot—intuitively. The results tell us exactly where to push next: tighter time-stamping and frame sync, better marker tracking or hybrid VIO/marker fusion, and latency compensation in the command loop. Those are straightforward engineering steps that close the gap while keeping the UX gains of AR. 

Congratulations on passing your Master's degree in Electrical and Computer Engineering

The topic of my final thesis was "Development and implementation of multisensory object recognition for obstacles at close range". I was motivated by the fact that in an ageing society, the number of people with physical disabilities is constantly increasing. Many of them are dependent on mobility aids such as rollators to maintain their independence in everyday life. A particular danger here is overlooking obstacles in the immediate vicinity, which can lead to dangerous falls. Preventing such falls is not only important from a health perspective, but also from an Economics point of view. By preventing falls, considerable costs can be saved in the healthcare system and the quality of life of those affected can be significantly improved. The topic of my master's thesis aims to extend an existing machine learning model based on YOLO and camera recognition. The aim is to investigate how a combination of stereo camera and one- or multi-dimensional LIDAR sensor can optimize object recognition at close range. This technology should be able to reliably detect various obstacles both indoors and outdoors and enable real-time reporting via MQTT.

We are thrilled to announce that Halyna, our talented student from Ukraine, has successfully defended her Bachelor's Thesis on "Methoden zur Erkennung von sicherheitsrelevanten Ereignissen in Zammad auf der Grundlage von Loganalysen vor dem Hintergrund des BSI Grundschutzes". This outstanding achievement is a testament to her hard work and dedication.

Despite the challenges posed by the war in Ukraine, Halyna has thrived in our program, demonstrating exceptional growth and resilience. With the support of her mentors and her own innate talent, she has overcome numerous obstacles and achieved outstanding academic success.

Halyna's experience and expertise will enable her to contribute to the rebuilding of Ukraine and foster German-Ukrainian cooperation in IT, economics, and science. We are proud to have her as part of our university community and wish her all the best for her future endeavors. 

Her success is a shining example of what can be achieved with determination and the right support.

We are pleased to announce that our student Yulia has successfully defended her Master's thesis in International Trade on the topic "Recent Marketing Strategies in the Cosmetics Industry: Integrating Artificial Intelligence, Social Media, Personalization, Leveraging Brand Ambassadors, Technological Advancements and Emerging Trends into Entrepreneurial Success". This achievement is a testament to Yulia's academic excellence, dedication, and commitment to her field of study. Throughout her academic journey, Yulia has demonstrated outstanding academic performance and professional skills. We are confident that her new role as a Content Marketing Manager will provide her with the opportunity to apply her knowledge and skills in a professional setting and make a lasting impact in her field. We congratulate Yulia on her achievement and wish her continued success in her future endeavors.