Highest Recognition for Danish Professor’s Quest for Better IT

New IT development needs to be more effective if technology is to keep pace with the need to manage the huge and increasingly complex systems important to society. With a rare Advanced Grant of 2.5 million euro from the European Research Council (ERC), Kim Guldstrand Larsen, Professor at Aalborg University’s Department of Computer Science, will now attack the problem in an entirely new way.

The coveted grant, which has never been awarded to a Danish computer scientist before, is the European Research Council’s highest recognition. The grant is given exclusively to “exceptional and established research leaders” who already have an excellent research track record and can take chances to break new ground in their field.

In Kim Guldstrand Larsen’s case, it is the reward for an innovative proposal on how the IT of the future can handle complex systems in the real world. The research project LASSO (Learning, Analysis, SynthesiS and Optimization of Cyber-Physical Systems) focuses on cyber-physical systems:

– This is about our eventually having IT everywhere controlling all possible processes and systems in the physical environment via sensors and embedded computers. But we have to do something to deal with the increasing complexity, says Kim Guldstrand Larsen.

Computer scientists at Aalborg University already work with companies in the energy and transport sectors on their specific problems with cyber-physical systems in planning public transport and coordinating energy consumption and energy production.

There we’re using the knowledge and the tools we already have. But in this new project, we have to get hold of basic techniques, algorithms and mathematical models so we can improve them and create a new generation of scalable tools, says Kim Guldstrand Larsen.

Models and machine intelligence

In practice, this will be done by combining two computer science disciplines, each with their advantages, in an unprecedented way to create a reasonable balance.

One discipline uses extremely thorough but therefore also complex and time-consuming models of reality to provide precise answers. This is known as model checking and is Kim Guldstrand Larsen’s area of expertise.

– We can model all parts of a system so we can say exactly what will happen and guarantee that it will work. The problem is that the models are enormously complex and therefore difficult to create. For example, we’ve put a great deal of energy into incorporating the time aspect so that things also work at the right time, he explains.

The other discipline uses a form of artificial intelligence to enable systems to make the best decisions possible from the available information and a probability calculation. This is known as machine learning and AAU’s Department of Computer Science has an entire research group in machine intelligence that will be part of the LASSO project.

Machine learning is used in many places throughout society where it is difficult to give absolute guarantees. For example, if you have a smart camera with auto mode, it tries to do as well as possible in a given situation. It’s not certain that it will get it 100 percent perfect every time, but you’re still glad that it gives you better quality images, explains Kim Guldstrand Larsen.

The idea of combining the hard guarantees of the models with the “as-good-as-possible” approach of machine intelligence gets praise from the European Research Council, which describes it as a bold but potentially groundbreaking idea. For Kim Guldstrand Larsen it makes a lot of sense to try:

– If you take a car, there are parts like brakes that absolutely have to work every time and on time. Alternatively, there may be other parts of the car’s systems, like the radio, where you can compromise on the 100 percent because it’s not vital. Similarly, the project is a confrontation between a technique that provides absolute guarantees by taking everything into account and another technique that optimizes what you can get out of the information available. It’s that combination that offers new opportunities to develop the IT we need in large cyber-physical systems, sums up Kim Guldstrand Larsen.

Important for Aalborg University as a whole

The ERC grant is a milestone in the professor’s own career, but the pride is also felt more broadly. Kristian G. Olesen, Head of the Department of Computer Science, had this to say:

– It is a remarkable distinction that only few achieve and a well-deserved highlight in an already spectacular scientific career that has been a pleasure to follow. At the same time, recognition at this level is important for the entire department and Aalborg University as a whole. It is something that gets noticed out in the world and it helps validate the quality of the research being done here.

Facts

  • The European Research Council (ERC) awards grants to researchers at various levels. An Advanced Grant is the largest and is given to “exceptional established research leaders so they can pursue ground-breaking, high-risk projects that open new directions in their respective research fields.” The competition was particularly stiff this year: The ERC received 2287 applications for a limited number of grants with success rate of around eight percent. Read more on ERC News.
  • The LASSO project (Learning, Analysis, SynthesiS and Optimization of Cyber-Physical Systems) will run five years with an ERC grant of 2.5 million euro that includes funding to employ a large number of PhD students and post-doctoral researchers.
  • The activities in LASSO will be coordinated with the activities in the Danish-Chinese center of excellence IDEA4CPS that also focuses on cyber-physical systems.
  • In addition to Professor Kim Guldstrand Larsen, the prime movers in LASSO are Associate Professors Radu Mardare and Manfred Jaeger of the Department of Computer Science at AAU, and Dr. Axel Legay of the research center INRIA/Rennes in France.
  • The ERC grant is the second in ICT research at Aalborg University in a short period of time. Earlier this year, Petar Popovski, Professor, Department of Electronic Systems, received a Consolidator Grant for research on effective and ultra-reliable network communication between machines.
  • In the EU’s 7th Framework Program for Research and Technological Development, AAU was the Danish university that brought in the most funding in ICT. AAU accounted for 35 percent of the ICT funding that Denmark received.

Further information

  • Kim Guldstrand Larsen, Professor, Department of Computer Science, AAU, Mobile: +45 2217 1159.
  • Kristian G. Olesen, Head of the Department of Computer Science, AAU, Mobile: +45 2567 5494.
  • Carsten Nielsen, Science Journalist, Aalborg University, Mobile: +45 2340 6554.

Source: AAU News