From the ops room.
Field notes, engineering logs, and the quarterly detection report. Written by the people who fly, drive and debug the machines — lightly edited, never inflated.
Why we fly blades at dawn
Field notesThermal contrast, wind budgets, and the physics of catching subsurface delamination before the sun flattens the signal.
2026-05-0241 defect classes, one honest confusion matrix
KVANT OSQuarterly detection report, Q1 2026: what we catch, what we miss, and why the corrosion-under-paint class still keeps the ML team humble.
2026-03-21The C-2 learns to cross weld seams
EngineeringEleven months of magnet geometry, told through the four prototypes that fell off a test tank so the fifth one wouldn’t.
2026-02-09BVLOS is paperwork, then it’s physics
RegulatoryWhat it actually takes to win a beyond-visual-line-of-sight corridor in Europe, and why we stopped resenting the SORA process.
2025-11-30Mission 041: teaching divers to love the M-1
Field notesRotterdam’s dive teams expected a rival. They got a scout that does the cold, blind, boring eighty percent — and hands them the dives that matter.
2025-09-14What a blade defect costs at each stage of denial
EconomicsFrom “it’s probably fine” to crane mobilisation: the cost curve of postponed findings, with numbers our clients let us publish.
2025-07-01Designing the 90-second battery swap
EngineeringA crew vessel in swell is a bad place for fiddly latches. How offshore gloves, not engineers, ended up designing the A-7’s battery bay.
We publish our error rates.
Four times a year, the full confusion matrix and what we’re doing about it. Ask for it with your demo.
Request demo