AI-Driven Steel Pellet Plant Optimization
Dataraft empowers Plant Heads to unify production data, predict bottlenecks, and optimize energy use-enabling faster, data-driven decisions and measurable cost savings.
Process Manufacturing Optimization: Steel Pellet Plant
Industry: Metals & Mining / Process Manufacturing / Pellet Plants
Users: Plant Heads / Production Managers / Process Engineers
Solution Area: Production Optimization, Predictive Analytics, Real-Time KPI Monitoring, Energy Management, Root Cause Analysis
Background & Problem Statement
A leading Pellet Plant operator in the metals and mining sector approached us with a critical challenge — their Plant Head was struggling with real-time oversight and decision-making due to scattered, lagging, and often siloed production data. Despite having automation in place (PLC/SCADA), the data lived in Excel sheets, control room systems, and standalone dashboards, making plant-level efficiency diagnosis and timely interventions nearly impossible.
We sat down with the Plant Head and key teams to understand how they were tracking specific energy consumption, pellet output vs. feed mix, kiln temperature variations, induration furnace performance, moisture levels, and equipment run-hours. A major concern was the reactive nature of their response to production bottlenecks, quality inconsistencies, and equipment inefficiencies, which were only identified after reports were compiled at the end of shifts or days.
🛠️ Our Approach with Dataraft
Greywiz deployed Dataraft to unify, clean, and contextualize data from various systems — SCADA, LIMS, historian systems, and operator logbooks. Here's how we structured the rollout:
Data Mapping & Connectivity:
We integrated Dataraft with their control systems to collect continuous data on critical process variables like kiln temperature, drum feed rate, gas consumption, and moisture levels in raw material and product.
Domain-Driven KPIs:
Working with the Plant Head and production engineers, we configured plant efficiency KPIs:
Energy consumed per ton of pellet
Thermal profile optimization across kilns
Pellet yield vs. planned output
Downtime tracking (planned vs. unplanned)
Dashboards & Alerts:
Custom dashboards were created for the Plant Head to:
Visualize trends across time periods (shift, day, week)
Drill down into deviations by line, equipment, or shift
Set up real-time alerts for critical thresholds
Decision Support:
Dataraft's AI layer enabled automated correlation analysis, highlighting, for example, how higher moisture in feed was impacting specific energy consumption, or how pre-heat zone inconsistencies were affecting induration performance.
🎯 Outcomes & Benefits
15% reduction in average energy consumption per ton of pellets over 6 weeks
Shift supervisors empowered with daily deviation reports and live dashboards
Faster RCA and corrective actions — from post-shift to near real-time
More confident, data-driven planning of raw material usage and kiln settings
Cross-functional collaboration between quality, production, and maintenance through a unified view of plant health
