Manufacturing & Industrial Vision// topic
Visual inspection and defect detection
In short
Whether a camera can actually judge your part: what defect classes are learnable, how few images you can start with, and what a false reject costs when the line does not stop for it.
9 pages
definitions
- False reject rate: what the line feels, not the accuracy on the slideA 1% false reject rate sounds like rounding. At 1,200 parts an hour it is 96 good parts a shift and over an hour of somebody re-checking them.definition6 min
- The anomaly score is a ranking; the threshold is a business decisionThe score orders parts by how far they sit from normal. It carries no units, no probability and no severity — which is why the threshold belongs to quality, not to engineering.definition6 min
- The golden sample is a decision record, not just a good partThe part in the drawer is the easy half. Its record — revision, attribute, approver, date, recheck trigger — is what an automated inspection inherits.definition6 min
- Writing the defect class list your inspection will be graded againstEvery class needs a name, a definition, a severity, a disposition and a minimum size. Classes two inspectors cannot separate produce a confusion matrix nobody can act on.definition6 min
diagnostics
- Half a part in frame: trigger and encoder faults that look like the modelStreaked surfaces, clipped parts and duplicated frames are acquisition faults with physical causes and arithmetic answers. Counting triggers against parts for 100 pieces tells you which one you have.diagnostic10 min
- It keeps up on the bench and falls behind at line speedA cell that is fast on the bench and late on the line has either a latency problem or a throughput problem, and they need opposite fixes. Queue depth over ten minutes tells you which, and a per-part budget tells you where the time went.diagnostic11 min
- Rejects double the morning a new coil or resin lot goes onA reject spike that starts at a lot changeover is a revalidation problem, not a retraining problem. The overlay takes an hour, and the 30-part retest tells you whether the material got worse or merely got different.diagnostic10 min
- The camera passed it and the customer found it: tracing an escape backwardsMost escape investigations start by retraining a model that never saw the surface in question. Working backwards through the retrieval chain — part to image to score to disposition — settles which of five very different faults you actually have.diagnostic11 min
- The model finds the scratch on day shift and misses it after darkA detection rate that tracks the clock is an optics fault wearing a model's clothes. Re-scoring yesterday's archived images tonight proves the model is unchanged and points the investigation at the light instead.diagnostic10 min
Other topics in Manufacturing & Industrial Vision
See allSafety, PPE and site monitoringCamera-based safety on a working site — PPE and restricted-zone detection, forklift and pedestrian conflicts, intrusion after hours — and the incident workflow that has to exist before any of it changes behaviour.7 pagesOn-premise, edge and plant data controlRunning models inside the fence: what hardware a camera count actually needs, how updates reach an air-gapped plant, and how to answer the question of whether any footage or production data leaves site.1 page
Working on something in this space?
Tell us where you are in a sentence or two. We'll tell you honestly whether we're the right team, and what a sensible first slice of the work looks like.
Start the conversation