Every DBX figure starts as an observation of a physical asset. This page explains how we turn observations into numbers, how we check them, and how we score ourselves against what actually happened.
DBX builds a digital twin of the physical commodity chain. Each link has its own signal and its own measurement method.
Satellite thermal observation of each plant gives an activity index from 0 to 100%. Coverage spans steel mills and copper, aluminium and nickel smelters worldwide, observed daily.
Stockpile volumes are measured from imagery and converted to tonnes using material-specific density. Loadings are matched to the vessels that take the cargo.
Vessel-level tracking follows each cargo from loading to discharge: origin, route, destination and estimated volume. Coverage spans around 95% of global dry bulk flows.
Discharges and receiving stockpiles show how much material is arriving and whether it is being drawn down or building up.
Thermal run-rates at the steel mills and smelters that consume the material show whether demand is holding up, plant by plant.
Thermal and optical satellite imagery of each asset, combined with vessel position signals across the world fleet.
Cloud, smoke and low-quality passes are filtered out. Readings that fail quality checks are flagged, never silently filled.
Each asset gets a physical measurement: heat at a smelter, volume at a stockpile, cargo on a vessel.
Measurements are tied to each asset’s known capacity and history, so a reading becomes a run-rate or a tonnage.
Asset readings roll up into country and global supply and demand balances, and AI models project them forward.
Commodity analysts check every release against market context before it is published.
Where operators and market participants hold their own records, we test our measurements against them before a dataset launches.
We will run DBX history against the series you trade and show you where it would have called the move.