The Challenge
A mid-sized integrated steel producer running an electric arc furnace and a hot rolling mill was watching its margins shrink. Power was its single largest cost, yet energy use varied widely from heat to heat. Unplanned breakdowns on the rolling mill stalled production for hours at a time, and surface defects on hot-rolled coils were often found only at final inspection, after the value had already been added.
Production planning ran on spreadsheets, disconnected from a legacy ERP. Nobody could see yield, energy or quality for a single heat in one place, so problems were discovered late and fixed by guesswork.
Binalyto’s Solution
We brought shop-floor automation, custom machine learning and ERP together into one connected operation:
- Furnace Energy Optimization Model: Machine learning that recommends the optimal power profile for every heat based on scrap mix, chemistry and furnace condition.
- Predictive Maintenance for the Rolling Mill: Vibration, temperature and load data from existing sensors feed models that flag failing rolls, bearings and gearboxes days in advance.
- Computer-Vision Surface Inspection: Cameras on the run-out table detect scale, cracks and edge defects in real time, so operators correct the process before an entire coil is lost.
- Automated Production Scheduling: A scheduling tool that sequences heats and rolling campaigns against the live order book, replacing two days of manual spreadsheet planning.
- Unified ERP with Heat-Level Traceability: Production, quality, maintenance and inventory connected in a single ERP, giving full traceability from the scrap yard to dispatch.
Delivered in phases, starting with a furnace energy pilot before scaling across the melt shop and rolling mill.
Key Results
- Annual Impact: ₹35Cr in combined savings and revenue uplift.
- Energy Efficiency: 18% less power consumed per ton of liquid steel.
- Reliability: 40% reduction in unplanned rolling-mill downtime.
- Quality: 30% fewer coils rejected for surface defects.
- Planning: Production schedules built in 2 hours instead of 2 days.
| Metric | Before | After | Improvement |
|---|---|---|---|
| Annual Impact | – | ₹35Cr | New |
| Energy per ton | Baseline | -18% | 18% |
| Unplanned downtime | Baseline | -40% | 40% |
| Surface defect rejections | Baseline | -30% | 30% |
| Planning cycle | 2 days | 2 hours | 90%+ faster |
Binalyto’s Proven Playbook
Automation, machine learning and ERP for heavy industry: melt smarter, roll faster, waste less.
“For the first time we can see every heat, every coil and every rupee in one place.”— VP Operations
