Adress: Office 1404, Al Thuraya Tower, Dubai Media City, Dubai, UAE
Email: artur@oesv.ae
LinkedIn: oesventures
Case study · Gold · West Africa
A cloud pilot in two weeks at a gold operation in West Africa
The first deployment outside the CIS, run on purpose in the conditions that break most pilots: ambient heat, dust, a rainy season that redraws the haul roads, and LTE coverage that disappears in parts of the pit.
- Operation
- Gold mine, West Africa
- Pilot scope
- Several haul trucks, one shovel
- Duration
- One month
- Conditions
- Heat, dust, rainy season, intermittent LTE
- Deployment
- Cloud, live in two weeks
- Outcome
- Full-scale deployment contract
Why this pilot mattered
Claims about cycle detection are easy to make in a temperate climate with reliable coverage. This operation offered the opposite: ambient heat that shortens hardware life, dust that coats everything, a rainy season that changes the haul roads week by week, and LTE coverage that disappears in parts of the pit.
The mine’s question was not whether the algorithm was clever. It was whether the data underneath it could be trusted well enough to act on.
What we did
A limited fleet — several haul trucks and one shovel — was instrumented and connected to a cloud instance. Work centres were geofenced, the road network mapped, and cycle detection run against live traffic for a month.
How we measured it
Detected haul cycles were compared against two independent sources: the operation’s manually recorded shift data and on-board records. The system recognised more than 95% of cycles across the pilot period. Deployment from contract to live operation in the cloud took two weeks.
The pilot led to a full-scale deployment contract, with a second stage covering auxiliary equipment, road condition monitoring and speed control.
What we cannot publish. The customer’s name, the exact country and the fleet size are not ours to publish — we do not hold permission for them, so we leave them out rather than guess.
We would rather describe the conditions precisely and leave the name out than describe neither.
Test us on your worst site, not your best
Connectivity gaps, dust and a mixed fleet are the normal case in this business. A pilot that succeeds on an easy site tells you very little.