Case study — Mining & ore

ArcelorMittal Mines

Following an alarm management study.

Mandate duration: 9 months
Scope: Pit fleet of hydraulic and electric shovels
Pillar engaged: Defect elimination & alarm management

Context

+3M$ in recurring gains from improved shovel availability driven by alarm analysis

Shovels set the pace of the pit: every hour of downtime translates immediately into tonnes not loaded. The units generated thousands of alarms per week, but nobody could say which ones actually announced a failure. Operators acknowledged by reflex, maintenance discovered the breakdown after the fact.

The problem

What we observed on the floor before the intervention.

Alarm flooding: several thousand events per week per shovel, with no priority hierarchy.
Systematic operator acknowledgement: genuinely critical alarms drowned in the noise.
No established link between alarm signatures and the failure modes observed on major components.
Recurring unplanned stops on hydraulic circuits, hoist motions and drives.
Alarm data not exported to the CMMS: impossible to trigger a notification from an early signal.

Our solution

The approach deployed with the crews, from diagnosis to autonomy.

01Full alarm management study: extraction, cleansing and classification of 24 months of fleet alarm history.
02Pareto analysis of the most frequent alarms, correlated with corrective work orders and production losses.
03Rationalization: nuisance alarms removed, duplicates merged, thresholds redefined and priorities set on three levels.
04Automatic CMMS notification rules created for failure-precursor signatures.
05Defect elimination on the root causes revealed by the analysis, with effectiveness verification.
06Operator and supervisor training on reading priority alarms and raising notifications.