Automation and Information Networks


Professor Hannu Koivisto

The research group of Automation and Information Networks (AIN) extends  the definition of traditional process and machine automation. Research targets areas from intelligent systems and neuro-fuzzy computation to applying modern telecommunication based remoting technologies in automation. Emphasize is on combining these issues, for example: system and control theoretic treatment of network management systems in telecommunication, using intelligent data analysis tools in automation and applying wireless telecommunication to automation.

The research goals are both on methodological and experimental. Application oriented development in co-operation with industrial partners has also been an successful topic. An important recent topic is participating research on information security of industrial automation and distributing the gained knowledge to Finnish companies.

Major Competence Areas

  • The system theoretic approach gives methodologies and tools for modelling, identifying and control system design of various dynamical systems, especially nonlinear models
  • Over 15 year competence on intelligent systems, neuro-fuzzy computation and intelligent data-analysis, both on methodological development and applications.
  • Developing and applying modern telecommunication based remoting technologies in various fields of automation.
     

Soft computing, modelling and identification

Soft computing research is targeted both in methodological development and application oriented research. Typical research projects have considered

  • Analysis of telecommunication networks and network traffic
  • Modelling and neuro-fuzzy control development of a vulcanization mixer
  • Modelling and operator support system development for a complex nickel leaching process
  • Bioaerosol detector development using hybrid genetic fuzzy systems
  • Supply chain control issues
  • Short term load forecasting system


Recent methodological result have considered

  • How to obtain small but still accurate fuzzy models and classifiers via multiobjective genetic search (Doctoral Thesis)
  • Develoment of a fuzzy probabilistic reasoning system and its application for identification of GARCH type models.


For more details, please see the  publications page.

Analysis of network traffic and telecommunication networks

The research considering network traffic analysis and corresponding management system development has been of major interest during last decade. The research has been mainly organised within joint co-operative research projects, funded by Tekes and industrial partners. Recently also some in-house research has been carried out.

  • End-to-end QoS analysis of a GPRS test network
  • Several network traffic analysis cases. Typical tools used: classical statistical measures, SOM-networks, Fuzzy clustering, Rough set tools
  • Quality of Service (QoS) requirements for a typical real-time automation network
  • Voice-over-IP QoS problem analysis with fuzzy probabilistic reasoning.


The published results can be found from the publications page: two doctoral theses, several M.Sc theses, journal and conference articles.

Supply chain control

The research considering supply control has been mainly organised within joint co-operative research projects, funded by Tekes and industrial partners. Major results:

  • Development of a modelling and simulation framework for supply chain control research.
  • System theoretic analysis and control system development, especially fuzzy and model predictive control.
  • Demand prediction subsystem based on fuzzy probabilistic reasoning
  • Optimal control studies for perishable item production and capacity
     

Applying telecommmunication to automation

Critical infrastructure and production protection is essential for enabling global competitivenes and securing Finnish innovation. The main objective of automation and information systems research group has been and is to make this possible by not only participating to state-of-the-art research in international forums but also by actively distributing this knowledge to Finnish companies.

Our applied telecom research concentrates on implementing security and studying real time requirements to all levels and to all distances in automation communication. On short distance communication our scope consists of analysis, evaluation and development of Ethernet based real time communication and implementing meaningful security for device level use. Our long distance topics are end-to-end quality of service for automation use and security requirements and challenges in global industrial systems communication.

We implement the distribution of research knowledge into practice by integrating security into all levels of course, writing Finnish publications from best practices view point and offering automation security courses for Finnish companies. Our research projects and theses always include security analysis as one important viewpoint. The automation and information networks research laboratory includes small scale industrial network consisting demilitarized zone, office network and two industrial control systems. This environment is used to give courses on securing control systems and revealing the ease of hacking unless the security is properly implemented and integrated into practices of automation personnel and subcontractors.


Updated by: Koivisto Hannu, 12.06.2011 20:36.
Content owner: Koivisto Hannu
Keywords: science and research
Top Feedback on this page