CORDIS - EU research results
CORDIS

Computational Intelligence Platform for Evolving and Robust Predictive Systems

Article Category

Article available in the following languages:

Robust software solution to optimise industrial operations

An EU initiative combined modern software, systems and knowledge engineering together with applications' experience of the process industry. This enabled the development of novel algorithms and a software platform for robust, adaptive predictive modelling applicable to different industries.

Digital Economy icon Digital Economy

The EU-funded INFER (Computational intelligence platform for evolving and robust predictive systems) project set out to develop a generic software platform that accommodates cutting-edge soft sensors and the latest adaptation and learning methods. Research focused on the domains of computational intelligence, software engineering and process industry/control engineering. INFER's main innovation was the creation and investigation of a novel type of environment in which the 'fittest' predictive model for whatever purpose would emerge – either autonomously or by user high-level goal-related assistance and feedback. The predictive system utilises any available feedback for its performance monitoring and adaptation. The INFER software platform, developed with the creation of evolving predictive systems in mind, supports parallel training, validation and execution of multiple predictive models, as well as an intelligent automation of the predictive model building process. A series of secondments enabled the exchange of know-how and experience for 16 industrial and 10 academic researchers. This resulted in contributions to over 70 leading publications and international conferences. One of the key applications examined was adaptive soft sensors for the process industry. A variety of application areas and contexts were used with the sensors. Project partners have acknowledged the growing importance of the interdisciplinary area of data science. A new Data Science Institute was established at Bournemouth University in the United Kingdom to further pursue INFER's long term-goals. INFER introduced an effective platform for predictive systems that is easily applicable not only to the process industry but to a variety of sectors. The software framework will contribute to transforming European industries from resource- to knowledge-intensive ones in order to maintain their leading market positions globally.

Keywords

Process industry, predictive modelling, computational intelligence, predictive systems, data science

Discover other articles in the same domain of application