LINDO API
by Lindo Systems
Steel Manufacturing
Optimising cuts for producing steel slabsIt's difficult enough to manage the cutting torches of a continuous caster that takes in molten metal at 2800 degrees and spits out red hot strands of solid steel slab. Adjusting the cut locations on the fly to minimise waste due to quality problems takes the already challenging problem to a whole new level.
Recognising this, MSA Process Automation Solutions and Services (MSA PASS) of Pittsburgh set to work to develop an application that would enable steel producers to automatically adjust the planned cuts as defects are detected by a quality tracking system. The result was their award winning Cut Optimisation application that reduces scrap and increases the yield of prime saleable product. The Cut Optimisation product calls the LINDO solver and utilises a proprietary mixed integer linear programming formulation. LINDO is able to typically solve the model in seconds – compared to alternatives tested by MSA PASS that required minutes.
Conservatively, yield improvements of 0.25 to 0.50% are possible. Given the huge quantities of steel that a single caster can produce in a year, this small improvement can translate into enormous savings. The Cut Optimisation software was selected as a winner of Industry Week magazine's annual Technology and Innovation Awards Program. The magazine's editors choose it as one of the Top 25 Technologies of the Year for 1999 that show high potential for redirecting business opportunities and creating new growth in the manufacturing world.
Another common problem steel makers face involves blending. Their challenge is to find the least expensive set of raw material ingredients that produce a steel of a specific final chemistry. Blend Optimiser, also developed by MPA PASS can help purchasing managers in planning the purchase of raw materials months in advance of their use. As with their Cut Optimisation application, Blending Optimisers uses LINDO to solve Mixed Integer Linear Programming models.