Why now
2026 is the year the industry got punished for not knowing.
The macro backdrop hands us the story. Margins are thin, the Street is punishing operational surprises, the workforce that used to carry site knowledge in its head is retiring, and regulators have stopped taking the operator's word for it. Here is why this matters now, and how Stratos goes beyond each of these requirements.
The Street stopped forgiving operational surprises
Five listed miners cut 2026 production guidance publicly. Several cited underground mining rates or site disruptions. Shares fell on the announcement in every case.
Company announcements and market reporting, 2026. Chart shapes are illustrative of the reported moves.
Stratos makes production claims video-verified, continuously, from the face to orbit. An operator running Stratos tells the Street what happened last shift, and shows the satellite pass that agrees with the mine plan.
Jan 22, 2026: shares fell over 10% after cutting copper and gold guidance on lower underground mining rates at Candelaria. Cut again Aug 13 after storms at Caserones.
Margin compression
The net profit margin of the world's top 40 mining companies was approximately 24% in 2011. It sits near 10% today. Critical mineral prices are at or near all-time highs over the same period. Record prices are not reaching the bottom line, so the margin is being lost inside the operation.
Statista, top 40 mining companies net profit margin series; PwC Mine and Deloitte Tracking the Trends
Stratos makes every efficiency decision verifiable. When margins are thin, you can't afford to act on guessed data. Every cycle, every delay, every minute recovered is traceable to footage. Efficiency becomes a measured outcome with proof.
Top 40 miner net profit margin
Commodity prices near record levels sit alongside margins that have fallen by more than half. Record revenue is not reaching the bottom line, so the margin is being lost inside the operation. Bars show approximate position against record highs.
Talent shortage
of mining leaders say talent shortages are holding them back from delivering production targets and strategic objectives. 86% say recruiting and retaining that talent is harder than it was two years earlier. Supervisors are promoted earlier with less experience, and the workforce that carried site knowledge in its head is retiring.
McKinsey Survey of Mining Senior Leaders and Executives
The camera becomes the experienced observer. The best crews' shifts become a replayable training library. Knowledge transfer happens through verified video, replacing the hope that a new supervisor remembers what the retiring one knew.
Downtime cost
Unplanned downtime on a large mining haul truck is commonly costed at $5,000 to $10,000 per hour, and published ranges run far higher once the disruption spreads through the cycle. Reactive repairs consistently cost multiples of the same work planned.
MapTrack equipment downtime statistics, 2026; industry maintenance cost studies
Stratos creates verified repair records. A repair is finished when the video says so. Wrench time, MTTR, and availability post from verified data, replacing estimates the board has to trust.
Regulatory pressure
The Global Industry Standard on Tailings Management requires monitoring, independent technical review and auditable records for high-consequence facilities. ISSB and CSRD reporting regimes are reshaping mining disclosure. Reclamation and disturbance claims are increasingly checked against independent satellite imagery rather than the operator's word.
GISTM standard text; ISSB and CSRD reporting requirements
Stratos builds the auditable record the ITRB wants. Ground video plus satellite reconciliation creates an evidence chain that's exportable for the ITRB, the auditor, and the regulator. Boards get verifiable operational data, replacing narratives.
Autonomy solved one phase. The cycle has seven.
Komatsu commissioned its 1,000th ultra-class autonomous haul truck in April 2026, at Nevada Gold Mines. That milestone covers one activity in one phase: haulage, mostly in large surface operations. Haul trucks are the single largest segment of the autonomous equipment market. Drilling, charging, ventilation re-entry, ground support and services still run with a person in the seat and a paper record, and that is where the shift is actually lost.
Komatsu press release, Apr 21 2026; Fortune Business Insights autonomous mining equipment market, 2025 segment share
Stratos verifies the whole cycle. Load haul dump is one phase of seven, and every phase is timed to the second from the same footage: face prep, drill, charge and blast, ventilate, load haul dump, ground support, services. The handovers between phases are captured too, so a delay has a phase, a duration and a video.
- Every phase of the cycle, not one activity
- Existing cameras, no retrofit and no autonomy project
- Phase-level cycle times with footage attached
- The waiting and the handovers counted, not estimated
The development cycle
Stratos verifies every phase, not one activity.
- Scale
- Wash down
- Mark up
- Tram in
- Set up
- Drill round
- Move off
- Load explosives
- Tie in
- Clear
- Fire
- Re-entry wait
- Gas clear
- Duct advance
- Spot
- Load
- Hang
- Queue
- Dump
- Bolt
- Screen
- Shotcrete
- Cure
- Air & water
- Power
- Track & road
- Survey
Autonomy instruments the truck. Telemetry instruments the engine. Stratos instruments the cycle: the waiting, the re-entry, the bolting, the services, and the handovers between phases where the shift is actually lost.
Vision AI is everywhere but shallow
BHP's FY2026 review credits computer vision on existing cameras for catching spillage and foreign objects. The market now accepts the input. Single point Vision AI solutions fail mining operations in three ways:
- Fragile: edge cases break models in dynamic underground environments
- Slow to train: mining requires unique models due to lack of general data sets
- Costly to re-train: new events demand excessive labour and cost to rebuild models
BHP, 2026; Becker, viAct, DeepX
Stratos trains AI models in 15 minutes. A single Stratos team member trains up to 200 models per week. Stratos never falls behind operations.
- 15-minute model training
- 200 models per week per team member
- Never falls behind operations
Contractor safety under the microscope
MSHA fatality reports published through 2026 repeatedly involve contractor personnel, across material handling, hoisting, mobile equipment and loading tasks. Contractors are under the microscope on every site. Owners assume the contractor is padding hours, and contractors have no record that proves otherwise.
MSHA published fatality reports, 2026
Stratos creates a record that protects the crew. The data defends your crew far more often than it criticises them. Every pre-start, every hazard card, every close-out is timestamped and attached to evidence. Contractors who run Stratos get paid on verification and walk into every claim with the replay.