https://www.mdu.se/

The iron and steel industry is compelled to innovate for survival amid escalating market competition, regulatory stringency, and elevated customer expectations. Significantly, the industry’s 7.2% contribution to global greenhouse gas emissions necessitates a sustainable and efficient approach. Central to this transformation is the adoption of advanced analytics, facilitated by leaps in computational technology, allowing real-time analysis of vast data sets and enabling valuable operational insights.

This study investigates the steel industry as a case study for legacy industries’ technology adoption, focusing on the integral role of Batch Data Analytics (BDA), Machine Learning (ML), and Cost-Sensitive Learning (CSL) as methodological cornerstones. BDA, through consolidation of batch processes, supports more efficient industrial operations. As a primary tool for extracting process data insights, ML fosters the develop-ment of focused datasets, while by connecting performance indicators to cost-oriented properties, CSL encourages a value-centric approach.

Legacy industries, including the iron and steel industry, present unique challenges in adapting to disruptive technologies, yet successful navigation provides abundant opportunities for innovation and growth. The introduction and assimilation of advanced analytics and disruptive technologies within established business models and processes are pivotal for these industries’ survival and growth. Given its significant economic role and continued reliance on fossil fuels, the steel industry exemplifies these challenges and opportunities.

The thesis details practical, real-world applications of these concepts, utilising case studies from Kanthal’s factory in Hallstahammar. These case studies highlight how advanced analytics can be applied to optimise processes, reduce costs, and improve productivity in actual industrial operations. They underscore the importance of contextualised data in batch processing settings and the need for improved data connectivity and interoperability.

Including intuitive and value-aligned Key Performance Indicators (KPIs), this research underscores the significance of a value-centric approach facilitated by CSL. The selected KPIs should accurately capture the processes’ value under investigation, aligning research objectives with the organisation’s operational goals.