Physical intelligence for commodity markets

We detect physical
supply disruptions before
the market prices them in.

QNT42 reads the physical world directly from satellite — refinery strikes in conflict zones, crop health across the global grain belt. Each disruption is flagged at the moment of acquisition, anchored to a satellite timestamp. Detection, not headline aggregation.

Median detection latency
0d
Ahead of public reporting
Yes
Live monitors
2
CONFLICT CRUDE  /  DETECTION PIPELINE
SAR Thermal NTL
Stage 01
Thermal signature
FRP60 MW
Stage 02
SAR backscatter
structural shift+5%
Stage 03
Disruption flag
anomaly+59.7σ

Coverage · 60+ refineries & chokepoints · global grain belt · Russia, the Gulf & key export regions

Conflict Crude Agricultural Health

Two monitors.
One physical read.

QNT42 runs two live detection monitors — supply disruption in conflict-zone refining, and crop health across the global grain belt. Each reads the physical world directly from satellite, against a hard historical baseline. Facts, not inferred scores.

MONITOR 01
Conflict Crude Intelligence
Physical detection of supply disruptions at refineries and export terminals in conflict theatres — Russia, the Gulf, key chokepoints. Sentinel-1 radar reads structural change; VIIRS thermal reads fire; night-lights read activity. A drone strike is flagged at acquisition — often the same day it happens, ahead of any public report.
SAR · Thermal · NTL 60+ assets Detected ahead of press
MONITOR 02
Agricultural Health Monitor
Satellite-derived crop vigour across the global grain belt — corn, soy and wheat in every major export region. MODIS NDVI measured against each region's 26-year seasonal normal, in robust units, surfaces stress (bullish) or vigour (bearish) 60–90 days ahead of official crop reports.
MODIS NDVI 26-yr baseline 60–90d lead
THE METHOD
Evidence, not inference
Every reading is a physical measurement anchored to a satellite acquisition timestamp — recorded at capture, never adjusted to fit later news. We surface what the sensors saw and the supply implication; we don't manufacture probabilities. That distinction is the product.
Timestamped Auditable track record No black boxes

Every strike,
timestamped at capture.

No headline aggregation. No black boxes. Each detection carries the satellite acquisition time, the physical signature, and — where it exists — the public confirmation that followed. The record is fixed at capture and never adjusted.

CONFLICT CRUDE
Recent detections
Sep 2026 · Russian refining strike campaign
detection log
timestamped at capture
SAR + thermal
Asset
Detected (UTC)
Signature
vs public report
Status
SlavyanskKrasnodar Krai
13 Sep 10:19
SAR +59.7σ
same day GUR / Kyiv Post
Confirmed
SaratovRosneft · Volga
11 Sep 00:07
16 hotspots
hours ahead ASTRA OSINT
Confirmed
AfipskyKrasnodar Krai
25 Aug 00:28
Fire on process
same day Reuters
Confirmed

How a desk reads this

Detection, independent of news
Each strike is flagged from the satellite pass itself — SAR structural change plus thermal fire on the process units — at the acquisition timestamp, before any report is filed.
Often ahead of the wire
Saratov was detected at 00:07 UTC; public OSINT confirmation followed hours later. Across the verified record, median detection latency is 0 days — many events flagged before the press.
Fixed at capture
The detection date is the acquisition date — never back-dated to match later reporting. That is what makes the track record auditable.
The signal is the physical detection. Corroboration, where it exists, is attached afterward — not required for the record. Detection log delivered via dashboard or REST.

Crop vigour,
against 26 years of history.

The second monitor reads crop health directly from satellite. MODIS NDVI for every major export region, measured against that region's own 26-year seasonal normal — surfacing stress or vigour 60–90 days ahead of official crop reports.

GLOBAL GRAIN BELT · NDVI
Crop vigour vs 26-year normal
Southern-hemisphere season · supply read surfaced both ways
MODIS 16-day NDVI
26-yr seasonal baseline
60–90d lead
Region
NDVI vs norm
Deviation
Read
PampaArgentina · wheat
+16%
+1.66σ
Vigour · bearish
Australia SENSW/VIC · wheat
+13%
+1.25σ
Vigour · bearish
US Corn BeltUnited States · soy
+2%
+0.67σ
At norm

A region reading well above its 26-year norm signals more expected supply — bearish; a region under stress signals the opposite. Because the baseline is each region's own history, the read is a physical measurement of crop condition, surfaced weeks before official yield estimates. NDVI is a vigour proxy, not a yield figure — the supply direction is what matters.

We see the event,
not the headline.

News wires and data aggregators report after an event surfaces publicly. QNT42 detects the physical event itself — from the satellite pass — and often before the wire. The difference shows up where it matters: timing, and coverage of the zones the press can't reach.

Capability
News / data aggregators
QNT42
Source of signalwhere the information comes from
Public reports, wires, filings
Direct satellite signal — SAR + thermal
Timingwhen you learn about the event
After it is reported publicly
At acquisition — often ahead of the press
Opaque zonesRussia, Iran, active war theatres
Limited — depends on local reporting
Full — the satellite sees what press can't
The recordintegrity of the event log
Narrative, revised as the story develops
Timestamped at capture · never back-dated
Basiswhat the read rests on
Aggregation and interpretation
Physical measurement vs historical baseline
Supply readfrom event to market implication
Left to the reader
Capacity affected, product, directional read
Inferencehow far the claim goes
Confident narratives, unquantified
Evidence surfaced · no manufactured probabilities

Built to survive
institutional scrutiny.

Detection is not a single hot pixel. Every disruption is established through a multi-step method designed to separate a real attack from routine flaring, attribute it to the right asset, and fix it in time — so the track record holds up under interrogation.

01  Multi-sensor detection
Radar, thermal & night-lights
A strike leaves three signatures: structural change in Sentinel-1 radar, a fire in VIIRS thermal, and a shift in night-time activity. We read all three, not a single sensor — so a detection is corroborated across independent physical channels before it is flagged.
SAR · VIIRS · NTL
02  Polygon attribution
The right fire, the right asset
Every thermal hotspot is attributed to an asset only if it falls inside that asset's mapped perimeter. This filters out neighbouring plants and agricultural burning — the difference between "a fire near a refinery" and "a fire on the process units".
Per-asset geographic perimeter
03  Own-baseline discrimination
Attack vs routine flaring
Each asset is measured against its own history, not a global threshold. A refinery that never flares and suddenly burns is an attack; a Gulf megacomplex that flares daily is not. The asset's own baseline is the discriminator that separates a strike from operations.
Per-asset historical baseline
04  Timestamped at capture
Fixed in time
A detection is recorded at the satellite acquisition timestamp and never back-dated to fit a later report. The event log is immutable at capture — which is precisely what makes the detection latency measurable and the record auditable.
Acquisition UTC · immutable
05  Evidence, not inference
We don't manufacture probabilities
We surface what the sensors measured and the supply implication — capacity affected, product, direction. We do not invent a "78% disruption probability" with no validated model behind it. The honesty of the read is part of the product.
Physical measurement only
06  Auditable track record
Measured, not asserted
Every detection is checked against the historical event record. Detection latency is measured against public reporting: across the verified strike record, median latency is 0 days — a meaningful share detected before the press. Claims you can verify, not marketing.
Median latency 0d · verified record

See the physical world
before it's priced in.

QNT42 is available to a limited number of institutional partners. Request access to receive the detection track record and methodology under NDA.

Limited availability · Institutional only · NDA on request

Talk to research, not sales.

Every access conversation is routed directly to our research team. We open with the detection track record, the methodology, and a live sample from the monitor that matters to you.

@
hi@qnt42.com
Direct to research desk
§
Detection track record
Shared under NDA
✓ Thank you — we will be in touch shortly.