Information Fusion from Databases, Sensors and Simulations
Contents
Executive summary
The University of Skövde proposes a research profile within the area of information fusion from databases, sensors and simulations, which entails the combination of data from multiple sources to generate information that cannot be derived from the individual sources. There are several reasons for focusing on this:
- The suggested area is unique, of strategic importance for industry, defense and public administration areas such as health care and is currently not represented in its entirety at any other Swedish university.
- A large number of industrial partners are participating in the development of the profile, which they consider an important area for future research and development.
- Two research platforms, Mechatronic Systems and Learning Systems, that were previously funded by the Knowledge Foundation, overlap this area. It is now a natural progression to build a strong and long-lasting research profile at the University of Skövde.
- The University of Skövde has a number of research groups working on different aspects of information fusion that allows development of a research profile of considerable strength, capable of cutting edge research on issues in information fusion.
- The profile is central to the main research focus of the University of Skövde and the application to the Swedish Government for the right to grant the Ph.D. degree. The establishment of a strong research profile in the important area of information fusion will be of great importance for the continued development of applied and basic research at the University of Skövde. This ensures that the profile will enjoy continued support from national and regional industrial partners, and from the University Board.
The vision is to build an internationally renowned profile that covers the entire spectrum, from sensors to decision support, i.e., from technical to human (cognitive and organizational) aspects. Such a profile will be of great interest to all organizations where decisions need to be based on different sources of information or where time is a critical factor and the decision support needs to be of highest possible quality, given the available information. Examples of the latter are when a company needs to act before a competitor does, or when a fighter pilot needs to act before his adversary does, i.e., when it is necessary to get inside the competitor’s decision loop.
Although there are several research groups at the University of Skövde already working on different aspects of information fusion, there is a large potential for further development and integration. The current research is internationally renowned within databases, real-time systems, artificial intelligence (especially learning systems), bioinformatics, information systems development, mechatronics, and cognitive science. The University is well established within these areas and has a critical mass of 9 professors, 12 researchers and 40 Ph.D. students. The aim is to further strengthen these groups and to develop new areas, e.g. uncertainty management. It is therefore of great interest to attract researchers who are excellent within the specific area of information fusion.
The profile entails a high degree of industrially related research. More than 10 companies have already expressed a sincere interest in the profile. Preliminary negotiations with these companies demonstrate that the profile will attract the necessary external funding to apply for the maximum level of funding from the Knowledge Foundation, i.e., MSEK 36 for the six-year period. We anticipate that the total budget for the profile, including funding from the Knowledge Foundation (MSEK 36), from participating industry (MSEK 36), from the University of Skövde and other sources (MSEK 28) will total about MSEK 100 for the six-year period 2005-2010. These figures have been used in planning the number of positions that can be filled by recruiting new research faculty and Ph.D. students into the research profile.
The nature of information fusion means that the scope of the research questions addressed by the profile is wide. The challenge is to further develop means for collaboration across the areas, research groups and companies. One industrial seminar has already taken place. It showed that very diverse companies do in fact have a genuine interest in the same central issues. One example is evaluation of the quality of the knowledge sources. This information is important in advertising for choosing the optimal media strategies, e.g., combining TV with newspaper advertisements. Within the network centric warfare initiative it is important to choose the optimal collection of sensors, e.g., for ground situation awareness.
Our strategy for maintaining and increasing industrial involvement in the information fusion research profile is based on three types of activities:
- Industrial seminars and workshops, providing knowledge transfer between the different profile partners, general dissemination of results, and opportunities to establish contact with additional industry partners.
- Project groups, in which researchers and a group of industries get together to discuss issues and form projects in specific application areas. One such group has already been formed within network centric warfare, with representatives from the University of Skövde, the Swedish Defense and defense industry.
- Joint projects, with active participation of both profile researchers and industry.
The development plan is to build on the existing research groups and complement or extend these with new professors or adjunct professors, researchers, Ph.D. students, and post docs. This ensures that there is a mix of personnel at different levels of research experience in the profile. The ambition is also to establish the first chair of Information Fusion in Sweden and to build a research group around it. A project organization with a director, a steering committee, a scientific committee and an executive committee will be established for all of this to be possible. The organization should promote cross-disciplinary projects and ensure that funds are allocated to projects that have the highest potential of contributing to the profile.
The expected results of the profile include:
- An increased ability to provide regional and national industry with competent personnel.
- The creation of at least three new businesses in co-operation with Gothia Science Park.
- At least 10 Ph.D. and 20 Masters Degrees within the area of Information Fusion.
- Significant research results directly applicable to co-operating industrial partners.
- Increased international academic recognition of the University of Skövde through research publications and collaboration at an international level.
Keywords: information fusion, decision support, simulation, and modeling.