Challenge details:

Steel scrap is charged in Electric Arc Furnace (EAF) for melting and converting to molten steel (refer Figure). Before charging the scrap in EAF, it is blended to achieve the desired scrap density and chemical composition of molten steel. The process of scrap blending for EAF includes analyzing the available scrap in terms of density and chemical composition. Properly blended scrap with good scrap density and scrap size can help in efficient furnace operation and high-quality steel production. 

Fluctuation in scrap density can result in un-melted scrap material which can stick to EAF sidewalls and thereby resulting in improper melting profile within the EAF. Currently, the supervisor observes scrap condition before charging in EAF. If the operator still finds unmelt scrap material sticking on sidewall, the operator will manually extend melting phase within the EAF which consume more power consumption to meltdown the scrap from sidewall. The scrap evaluation is carried out when the fresh scrap is being received & it is done by skilled operator using manual visual observation and sometime rechecking chemical composition by portable XRF. The bulky density is manually evaluated.

We are looking for a robust automated inspection system which can detect scrap density in real time from an unstructured heap of scrap present in the yard.