Full emissions control, from pit to processing plant.

Carbonova connects fleet, fuel-dispensing, energy-meter and processing-plant SCADA data to CarbonOS, a platform of AI agents that closes the Scope 1, 2 and 3 inventory, flags anomalies such as rising haul-truck idle time and ranks measures by cost per tonne of CO₂ avoided. The same dataset produces carbon reports for regulators, customers and investors, with an audit trail back to the source.

The two largest sources are usually diesel burned by haul trucks, loaders and generators and the electricity used for crushing and grinding, followed by explosives and, at some mines, methane. In a US Department of Energy reference study, grinding accounted for about 40% of the energy used in US mining and materials handling for 17% (DOE, 2007).

The EU Emissions Trading System covers metal ore roasting, sintering and pelletisation and any site with combustion plant above 20 MW of rated thermal input. Under the CSRD as amended by the Omnibus I directive, companies with more than 1,000 employees and more than €450 million in turnover report from financial year 2027, using the revised ESRS, which apply to financial years starting on or after 1 January 2027.

Each tonne of CO₂ from an installation in the EU ETS needs one allowance, and EU ETS auction prices averaged €82.32 per tonne in the third quarter of 2026. Imported iron ore pellets have carried a CBAM certificate obligation since 1 January 2026, which narrows the cost gap with producers outside the EU. Carbonova’s planning agent ranks fuel, power and process measures by cost per tonne avoided on the site’s real data.

Sentinel AI tracks fuel burn, idle time and load factor for each truck and flags drift, while the planning agent compares measures such as idle reduction, payload management, partial electrification and renewable power by cost per tonne of CO₂ avoided. Battery-electric haul trucks are still in field trials, so full fleet electrification is modelled as a long-term scenario rather than a baseline assumption.

Signal Ingest connects SCADA, ERP systems such as SAP or Oracle, energy meters, fleet IoT sensors and fuel and explosives purchase records, then normalises and versions the data. Proof Engine keeps each number’s lineage back to its source, with versioned emission factors and a change history, in line with the ISO 14064-3 verification standard, which makes third-party verification of the inventory easier.