ΑΙhub.org
 

New voices in AI: Tanja Kaiser


by
08 March 2022



share this:

Welcome to the third episode of New voices in AI!

This episode features Tanja Kaiser sharing her journey to working with swarm robotics. All episodes of New voices in AI are available here here

Who are you? Where are you from/ where do you work?

I’m Tanja Katharina Kaiser and I’m currently a research assistant and final year doctoral candidate in the Service Robotics Group led by Prof. Dr.-Ing. Heiko Hamann at the Universität zu Lübeck, Germany. My current research focus is on evolutionary swarm robotics, but I am interested in swarm intelligence and bio-inspired robotics in general. I hold a bachelor’s degree in applied computer science from the Baden-Wuerttemberg Cooperative State University Stuttgart, Germany, and a master’s degree in computer science from the Technische Universität Berlin, Germany.

How did you get into AI?

Back in my school days, I didn’t imagine becoming a roboticist or an AI researcher. But I’ve always been interested in technology and a variety of other things. I joined my school’s multimedia and science clubs, and my favorite subjects were physics and literature. When choosing my course of studies, I wanted to do something more applied than pure physics and ended up studying computer science. During my bachelor’s studies, I found the most exciting lectures were about technological approaches that are inspired by other fields of research, such as biology or psychology. Additionally, I developed an interest in robotics, which wasn’t in my curriculum. So I decided to focus on intelligent systems, including AI, in my master’s program. And from there, it was only a small step to becoming a swarm robotics researcher.

What are you working on currently?

My current research focuses on automatically generating controllers for robot swarms. The manual implementation of such controllers is challenging because it requires guessing what behavior of individual swarm members will lead to the desired behavior of the swarm as a whole. Automatically generating the swarm robot controllers avoids this challenge, but a suitable reward to drive the optimization process has to be defined.

In my research, I explore the minimize surprise approach in detail, which was first published by my doctoral supervisor Prof. Dr.-Ing. Heiko Hamann in 2014. We optimize controllers using simple evolutionary algorithms. But instead of rewarding a specific task, we reward that robots are good at predicting the state of their environment in the next time step. Robots have two options to improve their predictions: they can improve their internal world model, or they can adjust their behavior to match their world model. This gives the evolutionary algorithm the freedom to come up with a variety of collective behaviors. Using our minimize surprise approach in simulation and with swarms of real robots, we were able to generate controllers for several swarm behaviors, such as aggregation, dispersion, self-assembly, collective construction, and collective decision-making.

Next to my main research, I am working with a small team of students on the implementation of ROS2swarm, a ROS 2 package for swarm behaviors. It is a modular and easily extensible framework that provides a library of ready-to-use swarm behaviors.

What excites you about your work?

Most exciting is that my research is very multifaceted, since it is greatly inspired by nature. Evolutionary algorithms mimic natural evolution, rewards sometimes draw on concepts from other research fields, such as neuroscience, and swarm robotics is inspired by the swarm behavior of animals, such as ants, bees, or birds. This allows me to learn a lot about topics that go beyond my background as a computer scientist and to constantly broaden my horizons.

What is it about your specific sub-field you find interesting or intriguing?

When we think of intelligence and solving complex tasks, humans usually come to mind. But we tend to overlook the tiny creatures around us, such as bees or ants, that can accomplish a lot as a collective. The fact that we can draw inspiration from such natural swarms to have groups of simple robots solve complex tasks is what I find most fascinating about my field of research.

What would you like to ask the next interviewee?

Which aspects of your daily life inspire your AI research?

Where can people find out more?

A complete list of my publications is on Google Scholar, and I post about my research and more on Twitter (@msminirobot).



tags:


Joe Daly Engagement Manager for AIhub
Joe Daly Engagement Manager for AIhub




            AIhub is supported by:



Related posts :



Deploying agentic AI: what worked, what broke, and what we learned

  15 Sep 2025
AI scientist and researcher Francis Osei investigates what happens when Agentic AI systems are used in real projects, where trust and reproducibility are not optional.

Memory traces in reinforcement learning

  12 Sep 2025
Onno writes about work presented at ICML 2025, introducing an alternative memory framework.

Apertus: a fully open, transparent, multilingual language model

  11 Sep 2025
EPFL, ETH Zurich and the Swiss National Supercomputing Centre (CSCS) released Apertus today, Switzerland’s first large-scale, open, multilingual language model.

Interview with Yezi Liu: Trustworthy and efficient machine learning

  10 Sep 2025
Read the latest interview in our series featuring the AAAI/SIGAI Doctoral Consortium participants.

Advanced AI models are not always better than simple ones

  09 Sep 2025
Researchers have developed Systema, a new tool to evaluate how well AI models work when predicting the effects of genetic perturbations.

The Machine Ethics podcast: Autonomy AI with Adir Ben-Yehuda

This episode Adir and Ben chat about AI automation for frontend web development, where human-machine interface could be going, allowing an LLM to optimism itself, job displacement, vibe coding and more.

Using generative AI, researchers design compounds that can kill drug-resistant bacteria

  05 Sep 2025
The team used two different AI approaches to design novel antibiotics, including one that showed promise against MRSA.

#IJCAI2025 distinguished paper: Combining MORL with restraining bolts to learn normative behaviour

and   04 Sep 2025
The authors introduce a framework for guiding reinforcement learning agents to comply with social, legal, and ethical norms.



 

AIhub is supported by:






 












©2025.05 - Association for the Understanding of Artificial Intelligence