CASE story

Electronics manufacturing and industrial automation

Edge AI og Machine Vision automatiserer kvalitetskontrol af printplader

Applikation: AI-baseret kvalitetskontrol af printplader (PCB) med Machine Vision og automatisk analyse af produktionsfejl.
Produkt: JHCTECH BRAV-7721

The Task

In modern electronics manufacturing, printed circuit boards (PCBs) are becoming increasingly advanced, with smaller components, higher component density, and greater demands for precision. This places increasing pressure on quality control processes, where even minor manufacturing defects can affect the performance and lifespan of the final product.

Automated Optical Inspection (AOI) systems are widely used to detect potential defects on PCBs. However, these systems can generate a high number of false positives, incorrectly identifying functional circuit boards as defective. As a result, operators must manually review large volumes of inspection images to determine whether an actual defect is present.

This process is both time-consuming and resource-intensive. Furthermore, manual inspection results can vary depending on operator experience, fatigue, and workload.

To address these challenges, one of JHCTECH’s customers specializing in industrial Machine Vision has developed an AI-powered solution that automates a significant portion of the manual PCB re-inspection process. Using intelligent image analysis, the system filters out false positives, improves quality assurance, and significantly reduces the need for manual inspection.

Edge AI and Machine Vision Automate PCB Quality Inspection

The Solution – Edge AI and Machine Vision

The solution combines existing Automated Optical Inspection (AOI) equipment with AI-powered image analysis, using an industrial Edge AI computer as the central processing unit.

When the AOI system detects a potential defect on a printed circuit board (PCB), the images are analyzed by a Deep Learning-based AI model. Trained to recognize manufacturing defects, the model can distinguish between genuine defects and false positives.

This allows the system to automatically filter out a large proportion of false positives, enabling operators to focus only on defects that require further inspection.

The result is a more efficient and consistent quality inspection process, where AI handles the initial image analysis while operators perform the final verification.

Image analysis is performed locally on the Edge AI computer, allowing large volumes of image data to be processed directly on the production floor without relying on cloud-based AI processing.

Edge AI and Machine Vision Automate PCB Quality Inspection

Product – Industrial Edge AI Computer with NVIDIA GPU

JHCTECH BRAV-7721

The core hardware platform in this solution is the JHCTECH BRAV-7721, an industrial Edge AI computer designed for demanding applications in Machine Vision, image processing, and local AI inference.
The computer supports 12th, 13th, and 14th Gen Intel Core processors and can be equipped with dedicated graphics cards for GPU-accelerated AI processing.

BRAV-7721-WP|MEC system

The solution includes the following hardware configuration:

  • Intel Core i7-13700 with 16 cores and 24 threads for image processing and parallel workloads.
  • NVIDIA GeForce RTX 4070 GPU for accelerating AI models and image analysis.
  • Up to 64 GB DDR5 RAM for handling and buffering image data.
  • PCIe expansion slots for graphics cards and additional network interface cards.
  • Connectivity for up to 8 industrial GigE Vision cameras via PoE network cards.
  • Industrial communication interfaces for integration with production equipment.

In addition to AI-powered image analysis, the computer also manages communication with inspection equipment, image data storage, and data exchange with production systems such as MES (Manufacturing Execution Systems).
This brings multiple functions together in a single industrial hardware platform that can be integrated directly into the production line.

Reduced manual inspection and increased efficiency

According to JHCTECH, the solution has been implemented by several PCB manufacturers, including those producing advanced HDI (High-Density Interconnect) boards, rigid-flex PCBs, and IC substrates.

Results

The reported results demonstrate how combining AI with industrial Edge Computing can significantly reduce the need for manual quality inspection.

Reduced from typically 3–6 to 1–2 operators per AOI unit

Area

Reported Results

False Positives 70–90% automatically filtered out
Manual Re-inspection Over 70% reduction in workload
Staffing Requirements
Quality Inspection More consistent inspection results using AI
Traceability Digital recording of inspection data
Return on Investment (ROI) Estimated payback period of 8–12 months

The figures are based on the implementations reported by JHCTECH and may vary depending on the specific production line, AI models, and integration requirements.

The combination of AI and industrial Edge Computing can significantly reduce the need for manual quality inspection.

Edge AI creates new opportunities in industrial applications

This case demonstrates how AI and Machine Vision can be integrated into existing production environments without necessarily replacing current inspection equipment.

By combining industrial computers with GPU acceleration, cameras, and purpose-built AI models, manufacturers can automate tasks that previously required extensive manual processing. 

The technology extends beyond electronics manufacturing. Similar solutions can be used for visual quality inspection, object recognition, sorting, production monitoring, and automated defect detection across a wide range of industrial applications.

BRAV-7721 funktioner

This case study is based on a customer solution published by JHCTECH, a manufacturer whose products are distributed by Nord Technology. The described implementation and reported results are based on information provided by JHCTECH.

The images used in this case study are largely AI-generated illustrations.

← CASE story – Intelligent energy management – ​​maritime

Case story – The parking industry→

More:

Edge AI

What is Edge Computing and what can it be used for?

Machine Vision

The modern approach to industrial automation

Do you need a customized solution?

The same technology can be adapted to a wide range of industries and applications requiring local data processing, secure communication, and reliable operation.

Our products can be customized to meet your specific project requirements.

Contact Nord Technology to discuss your project or any technical questions. We’re happy to help you find the right solution.

Call us at +45 7027 0415.

Nord Technology - customized solutions

We are a small but efficient unit in our production team