XtalPi and Fangda Carbon Deploy AI to Refine Electrode Production

Graphite electrodes serve as the backbone of electric arc furnace steelmaking, requiring precise electrical conductivity and mechanical strength. Because raw material quality fluctuates by batch and supplier, manufacturers face constant pressure to adjust blends while navigating volatile commodity prices. Previously, refining these formulations relied on exhaustive, slow, and expensive physical testing. By integrating Fangda Carbon’s 60 years of proprietary production data with XtalPi’s data engineering, the companies have created a hybrid system that embeds expert rules into a predictive framework.

The model identifies optimal blend proportions that satisfy quality standards while driving down production costs. When supply chains shift, the system evaluates alternative inputs, allowing procurement teams to pivot rapidly. Validation confirmed the tool meets key performance indicators with speeds matching industrial production cycles. While human experts retain the final decision-making authority, the "compute first, verify later" approach significantly narrows the search space for viable material candidates. This deployment, emerging from a 2025 strategic agreement, marks the first core module of a broader project intended to eventually encompass comprehensive process design and advanced carbon materials like graphene.