The Regulator Already Has the Data Before the Mining Company Does

Brazilian mineral production totaled R$270.8 billion in 2024 and accounted for 47% of the trade balance surplus, volume spread across hundreds of active concessions, over 910 tailings dams, and land portfolios scattered across biomes with completely distinct climate regimes and regulatory requirements.

In April 2026, ANM established the Remote Monitoring Enforcement Policy and made it legally valid to issue an infraction notice based exclusively on geospatial data, with no on-site inspection. The regulator stopped depending on the mining company’s declaration and started cross-checking its own data against what the company reports, and ANM Resolution No. 220/2025 deepened this shift by broadening the concept of affected area to any stretch where tailings or pollutants could reach a protected area or disrupt economic activity.

An irregularity that used to become a liability only once an inspector reached the field can now be logged in ANM’s system before the company itself knows it happened. The question a mining company needs to answer has changed: it’s no longer whether continuous monitoring is necessary, but whether what it operates today works at the same information level as the regulator.

Why periodic inspection no longer closes the loop

Field inspection and sample flyovers cover what was inspected on the date of the inspection, while the regulator already operates with continuous coverage, cross-checking satellite data against what the company reported to verify whether the asset’s state matches the declaration, with independent evidence referenced to the same date.

In a heterogeneous portfolio, this problem gets worse, because dams, PRADs, operational perimeters, and land under licensing carry distinct risk logic and regulatory requirements, and applying the same inspection protocol to all of them produces insufficient coverage where risk is highest and unnecessary effort where the event is already expected. The two events that most often turn into liabilities before any field visit, wildfires and illegal construction, share a trait: a short progression window and a regulatory consequence that worsens quickly over time.

Wildfires: the gap between ignition and alert is the data that matters

A fire hotspot near a tailings dam doesn’t need to reach the structure to trigger a regulatory consequence. It’s enough to advance onto the surrounding environmental protection strip for ANM to log an occurrence in the direct-influence area, to reach a PRAD area under recovery and destroy the biomass the company had been documenting as evidence of meeting its license commitment, or to reach native vegetation within the concession perimeter and open the door to an IBAMA infraction notice, regardless of the fire’s origin.

Satellite detection covers this progression in layers: a thermal sensor picks up the hotspot’s radiometric signature with a revisit of just a few hours, before ignition crosses the threshold for critical expansion, while optical imaging validates the outline and confirms progression, and automatic prioritization by proximity to the critical asset decides which hotspot gets attention first, with no reliance on manual triage. The alert arrives with polygon, coordinates, distance to the nearest asset, and risk classification — never as a generic blob on a map someone still has to interpret before dispatching the field team.

At a mining operation now in its second year of continuous production with NOR, this design has sustained the detection of hundreds of fire hotspots before they became environmental liabilities, across a portfolio of over a thousand assets and tens of thousands of hectares.

Illegal construction: the early stage is the only window for effective response

A structure that starts with earthmoving can have its foundation finished within days and a wall up in just over a week. By the time periodic inspection finally reaches the stretch, what was negotiable before any investment is now an occupied building, and removal comes to involve legal proceedings, an indefinite timeline, and incomparably higher cost.

Satellite detection, in this case, works through multitemporal comparison: the system keeps a reference state for every polygon in the portfolio and automatically compares it against the current image, which makes small-area earthmoving, unregistered access openings, and surface-texture changes consistent with the start of a foundation all detectable with sub-meter imagery before the structure is even visible as a building. What separates early detection from a false positive, at this point, is resolution: with a 10-meter pixel, an early-stage structure a few meters across doesn’t even fill two full pixels, while with a 50-centimeter pixel the same structure occupies hundreds of pixels with geometry already distinguishable from its surroundings — and it’s that choice of source that determines whether the event is detected within the window where a response is still possible or after that window has closed.

Beyond early detection, the construction’s time series carries direct procedural value. When the company needs to show that the occupation is illegal and occurred after the concession began, the sequence of images with verifiable dates reconstructs the event’s timeline with independent evidence, which changes the company’s position in a repossession proceeding and in an ANM notification.

Why tropical Brazil requires a combination of sources

Persistent cloud cover can make the optical series unusable for weeks on end in the North and Center-West during the rainy season, exactly when hydrological risk to structures is highest and when fire hotspots in transition areas and illegal occupations tend to cluster. SAR penetrates clouds, smoke, and darkness and picks up incipient landslides, slope subsidence, and surface changes even under dense vegetation cover, complementing thermal and optical sensors in detecting and tracking wildfire progression and sub-meter imagery in the geometric detail that distinguishes illegal construction from noise. No single sensor solves tropical Brazil across every season of the year, and it’s this combination, not an isolated source, that keeps the evidence series continuous.

What the evidence chain needs to contain to have value before the regulator

Quality satellite data isn’t evidence on its own. A chain defensible before ANM and IBAMA needs to bring together source identification with capture date and time, versioned processing methodology, a cryptographic hash guaranteeing data integrity from capture to the final document, coordinates and a polygon referenced to the correct projection system, and event classification with documented parameters. Without these elements together, the data remains useful operational information, but not evidence with procedural weight — a difference felt precisely when the notification arrives, when a TAC is being negotiated, or when an infraction notice needs to be contested.

How NOR runs this cycle across a mining portfolio

NOR Space Intelligence closes the full loop, from the orbital image to the field ticket with an intact evidence chain, within the window where the decision still matters: multi-source coverage calibrated by asset class, a georeferenced alert validated before the ticket, evidence compatible with ANM and IBAMA, integration with the corporate GIS via documented API, with no dependence on a foreign vendor. At the mining operation mentioned earlier, this cycle has sustained monitoring of a portfolio of over a thousand land assets and tens of thousands of hectares, with every alert arriving already with a polygon, coordinates, operational context, and risk classification — never as raw data someone still has to interpret before acting.