KKAI research coming out of Klagenfurt regularly features at the world’s most prestigious artificial intelligence conferences. Patrick Rodler, from the Department of Artificial Intelligence und Cybersecurity, is one of the researchers frequently showcasing his work at these elite events. We spoke to him about his experiences, the research findings he has presented, and what his current projects mean for the future of the field.
Putting Klagenfurt on the global AI map
When people think of the heavyweight names in global AI research, they think of MIT, Stanford, Oxford, or the tech giants of Silicon Valley. It is by no means guaranteed that a Central European institution like the University of Klagenfurt can consistently compete in this arena. Yet, that is precisely what it continues to do.
Patrick Rodler, a researcher at the university’s Department of Artificial Intelligence and Cybersecurity (AICS), is one of the scientists leading this charge. He regularly presents his findings at the world’s largest and most influential AI conferences. This year, he once again attended the AAAI Conference on Artificial Intelligence in Singapore, where thousands of the world’s leading AI experts gather to debate pioneering work and shape the future direction of the discipline..
‘I am passionate about my field’
What is it like to represent the University of Klagenfurt at world-class events like these?
“It’s a special feeling every time – and a great responsibility, especially as I am often the university’s sole representative at these events,” Rodler says. “I’m incredibly proud to show that research from Klagenfurt can hold its own on the major international stage.”
Rodler is particularly proud of an exceptional recent milestone: within a single year, he had papers accepted at all three of the world’s leading AI conferences – AAAI, IJCAI, and ECAI. What makes this achievement even more remarkable is that he conceived, conducted, and wrote all three papers independently, without co-authors. “Getting there was far from easy,” he admits. “In my early days as a researcher, getting accepted just once felt like a minor miracle because the selection process is so fiercely competitive. But I am passionate about my field and I love my work. That helped me stay focused during difficult phases, keep going, and always put the quality of my research centre stage. I’m truly grateful that this work allows me to make a contribution to progress in such an exciting and vital field.”
More than merely presenting: connecting, inspiring, discovering
Conferences like these are far more than just a platform to present finished results. They set the pace for the entire discipline – they are environments where new concepts are born before possibly going on to change the world.
“Being immersed in these high-energy environments, surrounded by thousands of bright minds and a genuine flow of ideas, continually motivates me to think bigger, push boundaries, and look at problems from entirely new perspectives,” Rodler explains. “You meet people working on the same questions using completely different approaches, or tackling entirely different problems with similar ideas. It’s not unusual for a brief conversation on the margins of a session to spark a collaboration that later develops into a major joint project.”
A case in point: following AAAI 2026, Rodler used the connections he made to organise a workshop aimed at deepening co-operation with a partner university in Udine, Italy. The clear vision behind the move is to further establish the University of Klagenfurt as a key cross-regional research hub.
The research: when AI thinks like a human detective
What exactly does your research entail?
“My core focus is the automatic diagnosis and correction of faults in complex systems,” Rodler says. “In other words: how can we use AI to identify where a system is failing and get it running again with minimal effort for both humans and machines? It might sound abstract, but it is highly practical. These methods can be applied to software, hardware, robots, drones, vehicles, circuits, knowledge bases, recommendation systems, spreadsheets, industrial planning, and much more.”
At AAAI 2026, Rodler presented a paper on optimal abstraction levels in AI-assisted software debugging. He and his co-authors demonstrated that more detail is not always better. AI that reasons at the right level of abstraction can solve complex fault-finding tasks up to 10,000 times faster, often without any loss of accuracy. This marks a major step towards AI that combines human-like cognitive strategies with the raw speed of computers to solve difficult tasks more intelligently and scalably.
Meanwhile, at an ECAI conference, he presented a paper that resolved a research question that had remained unanswered for more than 35 years — first raised in 1987 by Raymond Reiter, a pioneer of AI research. Thanks to Rodler’s work, Reiter’s influential algorithm, which has been cited more than 4,800 times, can now be used efficiently in dynamic, interactive scenarios, offering speedups of up to 800% for users. His other work has focused on memory-efficient diagnostic methods that run reliably on low-resource hardware – such as mobile devices or Internet of Things (IoT) tech – where conventional methods fall short. Further research addressed the critical “measurement recommendation problem”, which previously acted as a major bottleneck in system diagnosis. Rodler’s solution can now point system technicians to the absolute best measurement and repair points within seconds.
Major projects, major impact
Rodler’s research does not just sit in an academic ivory tower. As the project leader of the FFG project SAELING – which commands a total budget of around EUR 1.8 million – ahe is working alongside colleagues in Klagenfurt, industry partners Siemens and voestalpine, and the renowned research department KU Leuven in Belgium to develop AI systems that optimise metal-cutting processes. The project bridges two strands of AI research that have largely been studied in isolation: rational AI (based on rule-based, logical reasoning) and learning AI (based on data-driven machine learning).
The expected real-world outcomes are substantial. “We aim to achieve energy savings on the scale of the consumption of entire small towns, alongside significant cuts to process costs through more efficient resource planning—all while substantially improving customer satisfaction,” Rodler notes. “Our interim results are already pointing very clearly in this direction.” The initiative is part of the AI for Green funding programme. Crucially, the prototypes being developed are designed to be generic so they can be adapted for other sectors, such as manufacturing plants or smart buildings.
At the same time, Rodler is a key researcher in the Bilateral AI Cluster of Excellence, a consortium of Austria’s leading AI institutions. Here too, the focus is on integrating rational and learning AI to make systems more versatile, autonomous, and broadly competent. “There is enormous potential for synergy when slow, precise, and provable logical thinking is combined with the fast, intuitive application of learned knowledge,” says Rodler. “This creates combinations of speed, rich information, clarity, and reliability that were previously hard to imagine. I believe we at the University of Klagenfurt are exceptionally well placed to make vital contributions to this area.”
What does the future hold?
Autonomous AI agents and LLMs: huge potential, many open questions
Another area currently dominating AI research is Large Language Models (LLMs) – the technology behind the chatbots that have become household names over the last few years. In industrial and commercial contexts, their full impact is still far from fully understood. These models display capabilities that continue to surprise even experts, particularly when they are not used in isolation but are combined and orchestrated as autonomous agents. For instance, they can now programme solutions to highly complex or previously unsolved problems at a level of quality that was virtually unimaginable until recently. Exploring these possibilities and understanding them scientifically is one of the most exciting current areas of research – and one that an AICS team, led by Gerhard Friedrich and including Patrick Rodler, is pursuing with immense energy. The team has developed a tool called “CheckMate“, which automates complex industrial optimisation problems and outperforms current market-leading systems. It will shortly be presented to the global specialist community at IJCAI 2026, one of the three top AI conferences.
However, the rise of LLMs inevitably brings new challenges. Because these models are prone to errors, research is increasingly focusing on the automatic detection and correction of mistakes in AI-generated outputs, the explainability of the results delivered, and proving the correctness of outputs – such as automatically generated code. “These will be the defining research questions in the years ahead – it’s a field with enormous potential,” Rodler says. It is a frontier that particularly fascinates him, given his years of expertise in exactly these areas.
.flex_column.av-qow95q1-b4f38848742cd5c188054fe06e6572dd{
border-radius:0px 0px 0px 0px;
padding:25px 25px 25px 25px;
}
#top .av-special-heading.av-lho0u7t-223cf1d74c0d8790043b64242bf77c36{
margin:0px 0px 0px 0px;
padding-bottom:0;
}
body .av-special-heading.av-lho0u7t-223cf1d74c0d8790043b64242bf77c36 .av-special-heading-tag .heading-char{
font-size:25px;
}
.av-special-heading.av-lho0u7t-223cf1d74c0d8790043b64242bf77c36 .av-subheading{
font-size:15px;
}
About Patrick Rodler
Patrick Rodler is a Senior Researcher and Lecturer (holding a Habilitation, the highest post-doctoral qualification) in computer science and artificial intelligence at the University of Klagenfurt. He has published more than 65 academic papers, including numerous single-author publications and papers in leading journals such as Artificial Intelligence and Information Sciences, and has presented his research at more than 40 international conferences. He also regularly serves as a programme committee member and reviewer for leading global conferences and journals. He currently leads the EUR 1.8 million FFG project SAELING and contributes to the Austria-wide Bilateral AI Cluster of Excellence as a key researcher.
His work has earned him the 2024 Research Award of the Province of Carinthia, multiple paper awards, an Outstanding Program Committee Member Award from the ECAI conference, and an award for teaching excellence from the University of Klagenfurt.
Der Beitrag Cutting-edge AI research: the University of Klagenfurt is a regular fixture on the global AI stage erschien zuerst auf University of Klagenfurt.