Solutions by industry / Metals & mining
Higher yield on every mining line, +$2M per line per year
The mining division of an international steel producer uses Wizata to find why identical lines yield differently, and to steer each one with real-time set-point recommendations.
The challenge
Identical lines, different yields
Some production lines delivered a lower yield in t/h than other, identically equipped lines. Yield optimisation had so far been done by non-digital means.
The producer wanted to know why, and to prove that a data-driven approach beats conventional methods on ROI. The causes were hidden in:
What the team set out to do
- Explain the yield gap between identically equipped lines.
- Hand yield optimisation to machine-learning models.
- Prove that digitalisation improves ROI over conventional means.
- Steer yield in real time to generate revenue.
Our approach
Three years of data, three levers, six lines
The Wizata platform aligned the process data with the business processes to show what really drives the production rate.
3 years of history
Three years of historical data were imported, cleaned, transformed and formatted for analysis, reporting and machine learning.
The three biggest levers
Re-aligned data showed what impacts the production rate most: mill feed, the sifting-screen speed, and the amount of water in the R spiral.
Real-time set points
Statistical and machine-learning models recommend optimal settings and actions to operators, steering yield in real time.
The outcome
More tonnes per hour, on six lines
Yearly increase per line, from a 1.5% average yield gain per line per year.
+1.5% yield
Average yield increase per line per year.
Six lines live
An AI recommendation system guides operators with optimal set points in real time.
Digital beats conventional
The data-driven approach delivered the ROI the producer set out to prove.
Built on the Wizata platform
Explore the platform →Next step
Find the tonnes your lines are missing
You don't have to do it alone. Talk to one of our engineers, or spend one or two days with our team on your own data.