Home Research Publications Events About LINK-SIC

Research within LINK-SIC

The area that LINK-SIC covers is, in broad terms Design of Systems for Control and Signal Processing based on Mathematical Models. A term that has been coined recently to capture the role of models and information for such design is Model-Based Information-Centric Systems Engineering. LINK-SIC will apply this perspective to problems of importance to Swedish industry.

Clearly, the topic is very broad and within a small effort only certain aspects can be treated. We will handle this by concentrating on specific areas selected by our industrial partners in the three nodes:
At the same time, the fascinating aspect of this research is that it is generic and generally applicable. A solution that has been generated for, say, an industrial robot can very well be applicable for an autonomous aerial vehicle. This makes it interesting and potentially rewarding to mix different application areas – in non-competing business segments – in a joint effort like LINK-SIC. It will be an important policy to actively pursue such synergies.

General Research Problems

The area contains several important research problems that will be treated in various aspects within the three nodes:

  • Design of Sensor Systems: Of particular importance is to integrate information of different types, such as radar-, vision-, map-, and inertial navigation signals. Another emerging sensor system problem is how to make a collection of small, cheap sensors communicate in order to create a more complete picture of the information state.
  • Supervision and diagnosis: Today’s more complex systems require advanced techniques to secure safe operation. To supervise the operation of a system and to diagnose when something is wrong, or when maintenance is required is of growing importance.
  • Autonomous capabilities: It is becoming more and more essential to have systems that take decisions on their own. It can be spectacular applications such as unmanned aerial vehicles (UAVs) or car collision avoidance systems, but it can also be automatic learning features in industrial robots or advanced speed control of vehicles.
  • Advanced control based on new complex models: There is an increasing need for more powerful mathematical models, like nonlinear system descriptions or hybrid models that cover both continuous dynamics and discrete events and logical constraints. (As a simple example, think of accelerator control and gear shifting in a truck.)