Distributed Industrial System Architectures : Exploiting Information and Knowledge to Increase Efficiency

Abstract

In industrial production, there are various environments where the potential efficiency is not reached. In process industry, the production processes include complexity and dynamics, which makes it difficult to the employees to operate them efficiently. Similar issues exist in mobile machinery. Especially inexperienced machine operators work in an inefficient manner, because it takes time to learn to control the machines. In addition, inefficiency occurs even in industrial business processes where information is exchanged. Related to process equipment, this exchange – especially between enterprises – lacks proper tools, which necessitates laborious and error-prone manual communication methods. Fortunately, information systems (ISs) can increase efficiency in all of these environments.

To increase efficiency, an approach is to design ISs that manage and utilise related knowledge in a computer-interpretable format. This thesis researches such systems in three environments. First, in the operation of machinery, the studied systems cover two purposes: performance-related feedback for operators and the adjustment of machine parameters. Both require knowledge management that is based on actual performance data. The second environment is industrial processes that receive benefit from operator assistance systems. Assistance eases demanding control tasks and especially plant-wide coordination. The third environment is industrial business collaboration, which lacks a digitalised medium for the exchange of technical information. Such a medium would apply knowledge about ecosystem-wide business practices. Although each of the three environments faces additional issues from distribution and asynchronism, the challenges are manageable with appropriately designed ISs.

As ISs are applied to improve efficiency, it is necessary to integrate heterogeneous information sources, which introduces challenges. Even if the information were in a computer-interpretable format, heterogeneity hampers communication. Heterogeneity occurs in various scopes, two of which are considered in this thesis. First, in production plants, ISs utilise heterogeneous, vendor-specific information formats. Common formats should be utilised instead to make systems integration more affordable to implement and maintain. Second, heterogeneity is even more remarkable in communication between enterprises. Enterprises have different practices, which reflects to their systems. Efficient collaboration necessitates the alignment of these differences, which is – fortunately – reachable with appropriate systems.

Considering efficiency, this work provides multiple results. The approach is to design suitable architectures for ISs. Regarding ISs, an architecture refers to the high-level structure and common principles that guide more detailed design tasks. An architecture is presented in each publication of this thesis. The publications are related to mobile machinery, production plants and the exchange of industrial technical information between enterprises. Concerning architectures, the publications provide multiple design challenges due to distribution, asynchronism and heterogeneity. In addition to architectures, this thesis introduces a conceptual model that covers the application of knowledge and the related information management tasks in industrial production. The suggested architectural solutions, as well as the conceptual model, are reusable in industrial production. Still, the results indicate that each problem requires careful design to conform to its specific requirements.