Expected Impact

In general, INFER is related to novel, intelligent techniques to process, analyse and classify huge volumes of data. The aim of the project is to develop a generic software platform accommodating autonomous self-adaptive, self-maintaining networked and embedded soft sensors as well as latest multi-level adaptation and learning techniques. The project will continue to utilise the most current research outcomes of machine learning techniques and transfer these to the industrial sector of manufacturing industries, which was initially proven to be a successful ongoing practice in a previous project. Establishing a new collaboration with the SME REC will equip the software with the modularity, robustness, scalability and user interfaces it needs to produce a successful commercial software product that can be disseminated to many other industries to enhance impact and practical applicability of the research outcomes. The participants of this project would not meet in their usual networks and by the continuous exchange of knowledge in terms of research results, practical applicability of techniques and future needs and requirements, a long-term collaboration will be established.

Impact on Industries

Due to increasing costs of labour, energy and resources in Europe, operational excellence is the key for competitiveness of Europe’s industry. Operational excellence leads to higher cost efficiency, better exploitation of plant capacity, loss reduction as well as achieving the compliance with environmental and safety regulations. INFER will provide a solution to such problems by researching and applying intelligent systems. Intelligent applications can deliver benefits on many levels and will support the process industries to improve their productivity and competitiveness, for example in improving maintenance, steadier low-level operation, decreasing costs of development and deployment of soft sensors, rationalising control room operations and boosting interoperability.

Relevance to SME

REC appreciates the benefits knowledge exchange with academia has on an SME and is already collaborating with both Wroclaw and Koszalin University of Technology in Poland. A new collaboration with BU will facilitate exchanges with academia on a European level and thus enhance the possibilities and range of knowledge to be gained. INFER will be the starting point of a long-lasting collaboration: During the runtime of the project, six REC-employees will have the opportunity to work in international teams in an academic environment for extended periods of time, thus being able to gain first-hand experience on producing and using high-quality research outputs in connection with commercial software engineering. The obtained knowledge will not only be of crucial value in the project proposed here, but will equip REC with expertise on machine learning and predictive systems. This consortium gives the SME an opportunity to develop products and services which, in their field of business activity, are the next generation of software systems and would gives them a competitive advantage and new potential customers. Being considered as technology front-runners also increases the possibilities and options of being regarded as sub-contractors to big companies in and beyond process industry. An increased attractiveness on the labour market for well-educated researchers and software engineers by providing a more diverse and challenging work environment will be a useful side effect. REC’s valuable input on designing postgraduate courses at BU will benefit the company in the long run by educating potential employees with both excellent academic as well as industrial expertise. Mutual visits for continuous exchange of knowledge through regular trainings and workshops will be beneficial and continue beyond the runtime of the INFER project.

Impact on academia

The applicability of research output to industrial problems is an ongoing concern and aim in academia. BU maintains several collaborations with industry in order to conduct research according to industrial needs, for example with British Telecom in the United Kingdom or Lufthansa Systems in Germany. In a previous project with Evonik industries, an exchange has been established leading to a successful development and proof of BU’s concepts and techniques to the process industry environment. On the way to a long-term collaboration and further joint research projects, this partnership will be extended in the INFER project with the inclusion of the software engineering company REC. By applying latest research outcomes to real-world problems, applicability of BU’s research will be proven, increasing its impact and strengthening BU’s reputation in a European and world-wide context. Opportunities arise for students in this collaboration which will include placements in the scope of a Master’s or PhD programme. Input of current industrial practices from both Evonik and REC will help to shape novel postgraduate courses like the MSc Advanced Computing and MSc Smart Systems and Technology to increase BU’s attractiveness and the number of postgraduate students at BU. Graduation of these highly qualified computational and engineering students will improve BU’s reputation and the ongoing collaboration will help to identify new challenging research programmes in the competitive market of the universities.


INFER aims at advancing techniques, algorithms, software, demonstrations and a systematic methodology of a new generation of soft sensors and accommodate them in a robust software framework. It will be an effective tool for control of large, medium and small scale industrial systems that will contribute to raising the machine intelligence quotient and to the transformation of the resource-based industries to knowledge-centered ones. The platform will be able to detect and analyse patterns, correlations, regularities and irregularities in process data and predict future behaviour resulting in universal techniques that gain knowledge from complex data and support decision making in many situations. High-quality software engineering will make the platform scalable, robust and applicable to a multitude of problems and support gaining access to new markets. To encourage dissemination of the actual product, software engineering expertise to ensure general interoperability and to set up appropriate customer-specific interfaces to the software platform is crucial. In this way, especially smaller and medium sized companies, who would normally not have the resources to develop their own internal solution, will be able to use the predictive platform. Dissemination of results within and beyond the runtime of the project will be further pushed forward by the involvement of REC, who will approach new customers and offer their services for customer-specific fine level adjustments to the software platform. Thus the project can contribute to the effort to transform European industries from resource-intensive to a knowledge-intensive sector in order to achieve and maintain leadership in the global market. Adding intelligence to devices and systems remains a key factor in obtaining a competitive edge in the EU. Some intelligent algorithms have been around since the 1950es and they have also been a popular topic in earlier EU FPs. The INFER project takes a novel approach to applying these methods: a combination of modern software, systems and knowledge engineering, together with application experience of the process industry. This will enable development of an effective platform for predictive systems that is easily applicable not only in process, but in many other industries.