IBM takes steel industry optimisation to new level: offers system to other industries

What’s more, IBM says its latest development, aimed specifically at melt shop optimisation, could easily be expanded to deal with other industries having complex manufacturing processes. It cites mining, chemicals and petroleum as likely beneficiaries. The system combines mathematical algorithms with steel industry knowledge to automatically derive the most effective way for steel producers to integrate real customer orders with dynamic production capacity across plant units with minimal waste. It’s a huge achievement. The solution, dubbed IBM Production Design & Operations Scheduling, is the first to encapsulate an end-to-end view of steel manufacturing – from the steel mill floor through all the process steps to delivery of finished products. It fills a yawning gap between a company’s ERP system and the plant’s real-time production control system by enabling optimisation of the daily plant-wide production process, typically over a seven-day horizon. While supply chain optimisation is a challenge for all industries, it is particularly complex in steel due to the sheer scale and number of interacting batch and semi-continuous processes. This system is said to handle everything from inventory management to slab, plate and cast design, the melting shop, hot strip mill and finishing line scheduling. IBM makes the point that optimising each of these areas independently (which is done routinely) results in savings, but that much greater gains can be achieved by going for the big one –simultaneously optimising all of them. “The combination of these scheduling issues with the volatility of market conditions makes production planning in the steel industry one of the most challenging problems facing manufacturers today,” says William Pulleyblank, vice president, IBM Centre for Business Optimisation. “By combining our consultants’ industry expertise with IBM Research’s mathematical modelling capabilities, this solution can help tackle these previously unsolvable problems.” That point is confirmed by early adopters, which say the system is helping them to improve production throughput, reduce inventories, cut operating costs and increase sales revenue. POSCO, for example, a major Korean steel manufacturer, is fulfilling orders more efficiently notably by packing multiple orders on a slab. The approach, which dramatically improves slab utilisation, was hitherto recognised as difficult to achieve from a computational perspective – but has been implemented effectively increasing average slab weight and reducing unused-slab weight, while increasing delivery performance. The latest enhancement to the system was developed at the IBM Solution Centre in Bangalore. It includes a new web user interface, support for multiple platforms (Linux, AIX, Windows), and extended capabilities.