Manufacturing AI quality inspection
Keep the existing AOI flow while inserting AI second-pass review into NG image handling, so humans focus on difficult samples.
AOI + AI inspection
Electronics manufacturing / PCB inspection · 某制造业企业 · 2026
The customer already used AOI optical inspection in PCB production. OK boards moved forward, while NG images were sent to the backend for manual review.
As line scale increased, manual judgment became inconsistent, review workload grew, and 24-hour production required costly three-shift staffing.
The customer wanted AI second-pass review for AOI NG images to reduce manual pressure and improve efficiency and consistency.
How Ouryun ships AOI + AI inspection for 某制造业企业
Ouryun introduced an Ascend-based AOI + AI inspection appliance, embedding trained AI models into the NG image review stage.
AOI flow integration
Keep the existing AOI first-pass flow: OK products continue, while NG images enter AI inspection.
AI second-pass review
A trained AI model reviews AOI NG images and allows AI-confirmed OK items to continue.
Human fallback for difficult samples
Samples still judged NG or uncertain by AI are routed to human review, reducing repetitive manual work.
Continuous model training
Inspection data, manual review outcomes, and inference results are collected for continuous model improvement.
Multi-line appliance deployment
The appliance provides compute, training, inference deployment, and line integration for 2-3 concurrent lines.
From connect to ship, in four steps
Connect AOI
Connect existing AOI equipment and backend image flow, defining OK / NG routing rules.
AI review
Send AOI NG images to the AI module for second-pass judgment and routing.
Human fallback
Route AI-confirmed NG samples to centralized human review for difficult cases.
Model iteration
Continuously update the model using manual review results and model monitoring metrics.
Customer profile
An AOI + AI inspection appliance for PCB production that performs AI-based second-pass review on AOI NG images and reduces manual review load.
Needs
The customer already used AOI optical inspection in PCB production. OK boards moved forward, while NG images were sent to the backend for manual review. As line scale increased, manual judgment became inconsistent, review workload grew, and 24-hour production required costly three-shift staffing. The customer wanted AI second-pass review for AOI NG images to reduce manual pressure and improve efficiency and consistency.
Solution
Ouryun introduced an Ascend-based AOI + AI inspection appliance, embedding trained AI models into the NG image review stage.
Impact
2-3 Concurrent lines supported; 24h Continuous production support; 降本 Reduced shift-based manual review cost; 复用 Reusable model capability for new lines
Numbers that prove value
Bring this capability into your business
More Ouryun case studies
Domestic AI inference resource pool
A unified domestic AI inference resource pool for a large manufacturing group, supporting multiple large models and group-level AI platform operations.
DeepSeek private compute foundation
A 6-node, 48-GPU private compute foundation for DeepSeek-R1 671B, including API publishing, AI security governance, and model security assessment.