Predavanje: Case-based reasoning applied in industrial intelligent systems

Sažetak

In computer science, Case Based Reasoning (CBR) is a method for implementing artificial intelligence algorithms, by identifying possible solutions to a problem using previously solved problems as guidance, inspired by the doctrine that similar problems have similar solutions. CBR proposes solutions to a new problem, based on its similarity to problems that have been solved before. If an accepted solution proves to be efficient, it is stored so that it can be used to solve future problems.

CBR has been used for various purposes especially in medicine and production industry. In industrial systems, CBR is particularly useful when an accurate model of the system is not available, usually because of a too complex relationship between controllable process variables and process outputs. In these situations, it is common that skilled technicians adjust processes when needed, by using their experience. The technicians’ experience may have been acquired through costly mistakes, whereas their memory is not as precise as a computer memory. CBR enables harvesting this experience and transferring it from experts to less experienced operators.

The lecture will introduce basic concepts of CBR, as well as possible synergies between CBR and statistical modeling methods like regression. Industrial assembly processes will be briefly introduced, and in particular the processes of assembling car body parts. The Cloud based architecture of the software system implementing CBR, for the optimization of a Volvo car crossmember body part assembly process, will be presented and demonstrated. The system was developed in a collaboration with Volvo Car Components, Sweden.

O predavaču

Ivan Tomašić has been a researcher and lecturer at Mälardalen University, Västerås, Sweden since 2013. Previously he was a researcher at Jožef Stefan Institute, Ljubljana, Slovenia, and Ruđer Bošković Institute, Zagreb, Croatia. Before that he worked as a research engineer and consultant for companies providing information-technology solutions. He received his PhD degree in computer science from the Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia, in 2013. His current work is focused on remote health monitoring of patients and elderly people; more specifically on monitoring and acquisition of physiological signals, as well as data processing and reasoning based on the acquired data. In addition to biomedical engineering (electrocardiography, cardiac electromagnetism, remote patients monitoring, biomedical signal processing), his research interests include storage and analysis of large and high-dimensional datasets, cloud computing data storage and analysis, applied statistical analysis, exploratory data analysis, as well as industrial intelligent systems.