Adopting Machine-Learning-Based Workflows for Reducing Production Risk and Cost

Transitioning production to machine learning (ML) workflows has many challenges. Pilots highlight the value of ML in isolation, but improving production requires active participation of plant operations. Easy-to-use software and contribution of key stakeholders can have an outsized impact on the long-term success of an implementation. Gerdau has been using Fero Labs’ ML software to reduce alloy addition costs and production risk in real time at its North American mills since 2020. This paper will explore pitfalls encountered during implementation, steps that were taken for adoption of ML software by key members of the plant operations, and the resulting benefits.

Pamir Ozbay | Fero Labs
 

Carter Almquist | Gerdau