Operational Reporting Platform for Pilbara Mining Operation

The Problem

A mid-tier mining operation in the Pilbara was running its daily operational reporting through a patchwork of spreadsheets, email chains, and manual data entry. Every morning, a team of three operations coordinators spent the first three hours of their shift pulling numbers from six different source systems, cross-referencing them in Excel, and compiling a report for the site leadership team.

The process was slow, but the bigger issue was trust. Numbers frequently didn’t reconcile. When the site manager asked “why is the crusher throughput figure different from yesterday’s report?”, the answer was usually a formula error or a missed data source. Decisions were being made on data that nobody fully trusted.

The client had attempted two previous reporting projects with larger consultancies. Both had stalled — one in a prolonged discovery phase, the other in a prototype that never reached production. The client was understandably sceptical of consultants promising reporting solutions.

The Approach

We started with a two-week scoping engagement — no code, just understanding. We mapped every data source, every manual step in the reporting process, and every question the leadership team actually asked of the data. The critical insight was that 80% of the daily report came from just three of the six source systems. The other three contributed edge-case data that was rarely consulted.

Rather than attempting to integrate all six systems simultaneously (the approach that had stalled previous projects), we proposed a staged delivery. Phase one would automate reporting from the three primary sources, covering the data that site leadership actually used for daily decisions. Phases two and three would bring in the remaining sources.

We delivered phase one in twelve weeks, with an additional two weeks of parallel running alongside the manual process to build confidence in the data. During parallel running, the automated figures matched the manually compiled figures within 0.3% — the discrepancy traced to a known rounding difference in one source system.

The Results

• Reporting time reduced from 3 hours to under 12 minutes — freeing shift supervisors for operational work.
• $47K annual saving — from eliminated manual effort and reduced rework.
• Data accuracy moved from “best guess” to auditable — with full lineage tracking and automated validation.
• Leadership confidence increased — decisions now based on data less than 15 minutes old, with clear provenance.

The platform has been running in production for over six months with zero unplanned downtime. The client has since expanded the platform to cover environmental compliance reporting.

“We’d tried this twice before with bigger firms and gotten nowhere. Charnwood scoped it honestly, delivered it in stages, and the numbers actually reconcile. That’s all we wanted.”