/answers/how-to-inspect-metal-castings-with-computer-vision/index.md · 42 lines
---
title: "How to inspect metal castings with computer vision"
description: "Photograph each casting at a fixed station after shot blast or cleaning, then use a model trained on the foundry own images to mark likely surface defects such as cracks, cold shuts, misruns, surface porosity, sand inclusions, flash left on the part and missing features. Each mark points to the part, pattern and spot on the surface, so a trained inspector can confirm whether the part passes, needs grinding or rework, or is scrapped. The model only looks at the surface it can see; it does not replace X-ray, pressure, dimensional or other tests the customer requires, and people make every disposition call."
url: https://lidvizion.ai/answers/how-to-inspect-metal-castings-with-computer-vision/
site: Lid Vizion
index: https://lidvizion.ai/llms.txt
---
[Direct answers](https://lidvizion.ai/answers/)
# How can a foundry use computer vision to check the surface of castings after cleaning?
Photograph each casting at a fixed station after shot blast or cleaning, then use a model trained on the foundry own images to mark likely surface defects such as cracks, cold shuts, misruns, surface porosity, sand inclusions, flash left on the part and missing features. Each mark points to the part, pattern and spot on the surface, so a trained inspector can confirm whether the part passes, needs grinding or rework, or is scrapped. The model only looks at the surface it can see; it does not replace X-ray, pressure, dimensional or other tests the customer requires, and people make every disposition call.
## What this means in practice
- Set up a fixed camera station and steady lighting after cleaning so each casting is shown the same way
- Collect images of castings that inspectors have already passed or rejected, for each pattern and alloy
- Label the surface defect types the quality team tracks, plus good examples of each part
- Send flagged castings to a trained inspector and log results by pattern, heat and shift
- Add reviewed images back into training whenever a new pattern, alloy or finish comes in
## How Lid Vizion approaches it
We start with the business decision and operating context, define ownership and measurable milestones, then use configuration, integrations, automation, or custom software only where each earns its place.
## More on Automation & custom AI
- [How can a wire harness or cable assembly shop use computer vision to check builds before they ship?](https://lidvizion.ai/answers/how-to-inspect-wire-harnesses-with-computer-vision/)
- [How can a plastics molder use computer vision to check parts coming off the press?](https://lidvizion.ai/answers/how-to-inspect-injection-molded-parts-with-computer-vision/)
- [How can a tile maker use computer vision to sort tiles for surface defects and shade?](https://lidvizion.ai/answers/how-to-inspect-ceramic-tiles-with-computer-vision/)
[Compare services](https://lidvizion.ai/services/) · [Use a planning tool](https://lidvizion.ai/tools/) · [Explore workflow use cases](https://lidvizion.ai/use-cases/) · [More direct answers](https://lidvizion.ai/answers/)
Talk it through
## Have a version of this question inside your company?
Bring the decision, the constraints, and what a good outcome looks like.
[Talk through the decision](https://lidvizion.ai/contact/?need=ai-workflow&from=%2Fanswers%2Fhow-to-inspect-metal-castings-with-computer-vision%2F#write) · [Book a call](https://calendar.app.google/bB26LdA9e2gmGFp69)