PCB Industry Solutions
The client is one of the top three mobile phone PCB manufacturers in Taiwan, targeting the high-end smartphone and tablet markets. With numerous parameters to monitor in PCB wet processes, existing equipment lacks smart sensing technology for real-time monitoring and dynamic adjustments. By deploying IoT and AI modules, the system enables real-time machine status monitoring, reducing anomalies and defective output.
AI Defect Recognition
Using Zero NG AI technology, ROI images are fed into an AI model for secondary inspection, dramatically reducing false alarm and escape rates.
About Zero NG
Lightweight AI model with cost-effectiveness
Only OK samples needed to train the first model
Quick training even from scratch
Easy for beginners with no AI background
AI Detection Success Cases
LED Defect Detection
Before
Overkill rate 1.3%, Escape rate 0.93%
After
Overkill rate 0.45%, Escape rate 0.07%
Die Bond Detection
Before
Traditional inspection with higher false alarm and escape rates
After
Overkill rate 0.15%, Escape rate 1.55%
Project Background
The client is a top-three mobile phone PCB manufacturer in Taiwan, specializing in premium smartphones and tablets. With complex parameter monitoring in wet processes, existing equipment lacked smart sensing for real-time monitoring and dynamic adjustments. IoT and AI modules were deployed to enable real-time machine status monitoring, reducing anomalies and defects.
PCB Electroplating Real-time Monitoring
IoT collects and monitors current, voltage, copper sulfate concentration and other parameters in real time, visualizing critical data on dashboards while continuously recording process variations and integrating with automatic dosing systems for dynamic parameter adjustment.
IoT real-time collection and monitoring of process parameters, combined with AI algorithms and statistical analysis
Copper sulfate concentration trends visible every five minutes — anomalies handled immediately
On-line SPC provides real-time electroplating thickness trends; Pareto charts reveal major defect causes
Management can monitor various parameters at any time for comprehensive production oversight
Key focus: Sensing → Visualization → Analysis & Feedback
PCB Electroplating Prediction Platform
An intelligent electroplating thickness prediction model trained on domain expert experience provides smart reference values for machine parameter settings, reducing experiments and accelerating yield improvement. AI models trained on operator expertise minimize test runs and shorten time to higher yields.
PCB Industry FAQ
Zero NG technology requires only OK samples to begin training, and can be completed quickly even from scratch.
Recommended Products
Manufacturing Execution System
DIAMES
- Digitized production process management
- End-to-end production tracking
Equipment Automation Platform
DIAEAP
- Equipment connectivity and IIoT
- Real-time data acquisition and OEE monitoring
Manufacturing Execution System
EXC-MES
- Central hub for smart manufacturing
- Real-time shop-floor visibility and control