MANUFACTURING
Predictive Maintenance
Predictive Maintenance. What it takes to run it.
Visual monitoring of equipment wear to schedule service.
// 01 · The question
“Which machine gets serviced at the next planned stop?”
// 02 · Why it is hard
The useful signal accumulates over weeks, so the dataset is a time series of the same equipment rather than a set of independent images, and it is worthless unless the frames are comparable.
// 03 · What the labels have to be
Instance segmentation
Decided before the first frame
The signal is progressive: belt wear, leak staining, thermal spread. All of them are judged by how much has changed since last month, which is an area comparison and needs an area label.
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
Defect Annotation
Pixel-level defect annotation for surface scratches, cracks, dents, weld faults, and assembly misalignments. Multi-sensor alignment for multi-camera inspection systems.
LexInsight
Line Performance
Connect defect detections to production data - batch numbers, machine IDs, shift schedules. Identify defect patterns across lines and predict maintenance needs before unplanned downtime.
LexAlert
Quality Alerts
Monitor defect detection accuracy across changing production conditions. Detect when lighting shifts, new materials, or equipment aging degrades model performance. Auto-retrain to maintain 99%+ accuracy.
// 05 · What breaks it after launch
A serviced or replaced machine resets the baseline it was being compared against. The trend line breaks, and unless the event is recorded the model reads a repair as a sudden improvement in condition.
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 manufacturing
Defect Detection at Scale
The Challenge
Vision models performed well in pilots but lost accuracy as factory conditions shifted.
Our Execution
Built pixel-level annotated datasets and multi-sensor alignment for reliable defect detection.
The Result
Maintained 99%+ defect detection accuracy across changing production environments. Reduced unplanned downtime.
99%+
Accuracy maintained in production