OIL & GAS
Hazard Zone Intrusion Detection
Hazard Zone Intrusion Detection. What it takes to run it.
Immediate alerts for unauthorized personnel in red zones.
// 01 · The question
“Did anyone enter the red zone, and when?”
// 02 · Why it is hard
The zone lives in image space, not in the world. Everything about this use case depends on the camera meaning today what it meant the day the zone was drawn.
// 03 · What the labels have to be
Bounding box
Decided before the first frame
A person box plus a zone drawn once in image coordinates. The whole decision is whether a point on the box crosses a line, so the geometry is as cheap as detection gets and the effort belongs in the zone definition instead.
This is the decision that is expensive to reverse. The geometry has to match what the answer contains, and finding out it does not means labelling the set a second time.
// 04 · How we run it
Three parts of one loop, on this job.
LexAnnotate
Precision Safety
Annotate rig images with bounding boxes for machinery, valves, flanges, and personnel with PPE classification. Safety-focused labels tailored to oil & gas operations. 12k+ hazardous-zone annotations in production deployments.
LexInsight
Corrosion Tracker
Link visual detections to asset management systems. Map corrosion to specific valve IDs, track progression across inspection cycles, connect to maintenance scheduling and compliance codes.
LexAlert
Safety Alerts
Monitor deployed safety models against environmental changes - weather, lighting, equipment repositioning. Detect drift before it causes a missed hazard. Route failures back for immediate retraining.
// 05 · What breaks it after launch
This is the use case camera drift ruins fastest and quietest. Nudge the mount a few degrees and the red zone now covers a patch of walkway. Detection accuracy is unchanged, every alert is wrong, and no model metric will show it.
It is not the only one that can get this use case, it is the one that usually gets it first. All ten conditions.
// 06 · In oil & gas
Hazardous Zone Monitoring
The Challenge
Lacked labeled rig imagery, delaying AI model training. Relied on manual monitoring in hazardous zones.
Our Execution
Annotated rig images with bounding boxes for machinery and personnel. Delivered precise, safety-focused labels.
The Result
Faster AI deployment with rapid model iteration. Improved safety through automated detection.
12k+
Precision image annotations delivered