Equipment Fault Detection and Classification System - Intelligent Error Detection Classification - Defect Classification Control - FDC System - Semiconductor FDC - GTRONTEC

Fault Detection and Classification (FDC)

Leverages process variable control and statistical analytics to detect and analyze equipment abnormalities in real time, enabling immediate corrective actions and improving Overall Equipment Effectiveness (OEE) and product yield.

Process Modeling | Data Collection | Summary Calculation | Detection Modeling | Data Analysis | Detection Simulation | Linkage Control

Pain Points

1. Old Technical Architecture

Monolithic architecture that lacks microservices support, high availability and dynamic scaling, resulting in poor performance and low system stability.

2. Incomplete Functionality

Limited data processing and computation capabilities, low flexibility, delayed real-time alerting and no support for custom control rules.

3. High Implementation Difficulty

Complex system deployment with no support for containerization, cumbersome initial setup and heavy reliance on custom development logic.

4. Poor Usability

C/S architecture that is cumbersome to use; requires managing numerous models, generates excessive false alarms and lacks complete data interfaces.

Core Advantages

  • New Architecture with Excellent Performance
    Built on a high-performance big data architecture, supporting high-throughput, low-latency real-time processing, highly available distributed deployment and microservices with containerization for elastic scalability.
  • Comprehensive and Complete Functionality
    Supports multiple data acquisition channels and flexible collection strategy management; enables window-based aggregation and custom scenario computations; fully compatible with West Electric control rules with support for custom extensions; delivers real-time, historical and simulation-based data analytics reports.
  • Good Product Usability
    Built on a B/S architecture with browser-based access; supports flexible deployment options, including on-premise localization; enables process-type modeling with one-click management of tens of thousands of instances; provides fully open interfaces for custom report development.
  • Convenient and Fast Operation and Implementation
    Provides guided, foolproof deployment with simple configuration; supports virtualized, containerized deployment with high availability; enables batch import/export for fast setup; and offers fully configurable system logic with no coding required.

Application Scenarios

Panel Manufacturing

Controls key process parameters across coating, etching, exposure and grinding to improve product yield.

Silicon Wafer Manufacturing

Controls process parameters across crystal pulling, wire cutting, grinding and polishing to reduce process-related incidents.

Wafer Manufacturing

Controls process parameters across grinding, polishing, epitaxy and related processes to reduce equipment failures.

Chip Manufacturing

Controls process parameters across coating, etching, exposure and grinding to reduce scrap rates.

Cases

Case: FDC Project for a Panel Enterprise Module Factory

Project Background: As the module factory’s new plant prepares for mass production, there is an urgent need to establish an advanced FDC system. The existing self-developed FDC platform lacks the functionality required to support new-generation equipment and control requirements. The project scales to manage over 3,000 machines, 100,000+ data points, and 200,000+ concurrent detection instances, while supporting simultaneous access for more than 1,000 users.

Project Results: Strengthened fail-safe monitoring for parameter registration and sensor miss detection, with improved usability for registration and data queries to enhance operational efficiency. Leveraging adaptive AutoRange control thresholds significantly reduces false alarms. Additionally, group-based equipment management aligned with production needs enables more precise sensor miss monitoring.

Case: FDC System for a 12-inch Chip Manufacturing Enterprise

Project Background: With a new plant under construction, the original FDC system vendor can no longer provide support due to policy constraints, creating an urgent need for a domestic FDC replacement solution. The project scales to manage over 1,000 machines, more than 1 million data points, and over 2 million concurrent detection instances, while supporting simultaneous access for more than 1,000 users.

Project Results: The FDC system collects data directly from equipment via virtual port software, supporting Interface A protocol and up to 10Hz sampling frequency. It accommodates diverse heterogeneous data sources with fully configurable parsing logic—no coding required. Comprehensive summary algorithms and control rules support customer-defined configurations and seamless system import. For large-scale data queries, the system delivers high-performance analytics, processing 30 days of data within 1 minute with instant chart visualization.

Case: Smart Manufacturing Integration Project for a Semiconductor Silicon Wafer Manufacturing Enterprise

Project Background: The existing production process suffers from weak process control, inconsistent product parameters, self-induced defects, and significant labor inefficiencies, creating an urgent need for improvement through an FDC system. The project covers key process equipment including sticking, wire cutting, degumming, and cleaning, managing over 200 machines, 10,000+ control data points, and more than 30,000 detection instances.

Project Results: Integrated three core modules - fault detection and classification, intelligent quality analysis, and process optimization - on a unified data platform to enable seamless data integration and sharing. Established a structured knowledge base of key problem-solving methods for reuse across factories. Within three months of deployment, the system delivered measurable results, increasing product yield by 5%, reducing man-hours by 10% and significantly improving production collaboration efficiency.

Functional Modules

Detection modeling of parameters, defining detection dimensions, data filtering, detection rules, setting detection thresholds;

Contact Us Online

Follow Us

Homepage Phone Contact us online
联系方式 +

*您关注的问题?

*您的联系方式?

*怎么称呼您?

*您的公司名字?