Blog · AI visual inspection
Notes from the inspection line.
Anomaly detection, few-shot training, edge deployment and the metrics that matter: practical writing for solution partners, quality managers and automation engineers.
- 11
Rule-based vs AI machine vision: when to use which
When to use rule-based machine vision, when to use AI inspection, and when to combine them. A decision framework with a side-by-side comparison for quality and automation engineers.
3 September 2026 · 7 min read
- 12
How to inspect a part with only 20 images
Modern AI inspection doesn't need thousands of images. How anomaly detection and few-shot learning inspect from about 20 good samples, and why lighting, not data, is the real bottleneck.
31 August 2026 · 7 min read
- 13
The inspection metric vendors don't advertise: false-negative rate
A 99% accuracy claim means little without the false-negative rate. The four inspection metrics that matter, why escapes are the expensive error, and the questions to ask any vision vendor.
27 August 2026 · 7 min read
- 14
Edge vs cloud: why line-speed AI stays on-device
Line-speed inspection can't wait on the network, and many plants keep raw imagery on site. Why edge AI leads for visual inspection, and where cloud still helps.
24 August 2026 · 7 min read
- 15
AI inspection in 30 minutes, without a vision engineer
A pre-configured inspection unit turns a multi-month vision project into a 30-minute field task. The four-step field guide, Mount, Aim, Configure, Wire, for integrators.
20 August 2026 · 6 min read
- 16
The cost of a missed defect: putting a number on one escape
What one escaped defect really costs downstream under the rule of ten, plus direct and indirect cost, and how a 0.69% false-negative rate keeps escapes from reaching the field.
17 August 2026 · 7 min read
- 17
Build vs buy a machine vision system: a decision framework for 2026
When to assemble a custom camera-lighting-PC-software stack versus deploy a pre-integrated unit, framed around engineering hours, not licence lists.
13 August 2026 · 6 min read
- 18
What is anomaly detection in manufacturing?
Anomaly detection learns what a good part looks like and flags anything that deviates, catching the rare, first-seen defects a fixed rule was never written for.
10 August 2026 · 7 min read
- 19
What is edge AI in visual inspection?
Edge AI runs the model on a device at the line, on fanless Jetson-class compute, deciding in about 30 ms with images staying on the line.
6 August 2026 · 6 min read
- 20
Machine vision vs computer vision vs AI vision: the terms, untangled
A factory-floor glossary of three terms buyers use interchangeably, computer vision, machine vision and AI vision, and where a hybrid edge-AI unit fits.
3 August 2026 · 6 min read