The path to continuous improvement for semiconductor manufacturers got a lot easier today with the announcement by Aidentyx, Inc. that it has rolled out the chip industry’s first Manufacturing AI framework featuring multiple applications, each backed by specialized AI engineering agents orchestrating smart, secure, and net-zero-friendly AI analytics that promote greater interoperability, collaboration, and increased performance among applications and data sources. According to company CEO, Jason Kim, “This first GenAI system will transform fab operations by providing deeper insights, predicting potential failures before they occur, optimizing maintenance, and quickly prescribing remedies to everyday challenges facing fab managers everywhere, especially as it pertains to unscheduled stoppages.”
The new system was developed to help empower engineers to deliver true, actionable insights and higher levels of support and performance across equipment, sub-systems, and processes. Each AI Agent leverages data insights from across the PODO® cross-application AI framework, which houses a trove of manufacturing-rich data, to orchestrate a highly differentiated set of data analytics applications that provide IDMs and OEMs with game-changing results. Specialized AI Agents work alongside cleanroom and sub fab engineering teams to deliver better, faster, fab-wide insights of issues that impact production processes, and provide up-to-the-second status on the health and performance of critical sub-systems and production assets.
The system was developed by the same innovators who delivered the industry’s first leading Fault Detection and Classification (FDC) system. “We are excited to offer engineers another industry first – a data analytics system with a broad range of equipment, sub-system and process analytics applications with AI Agents to better connect and align sub-fab and cleanroom operations to deliver millions of dollars in production value, while dramatically reducing overall fab maintenance costs,” said James Na, Chief Product Officer, Aidentyx. “For the first time, our industry has a system that offers engineers a holistic view of the health and performance of critical assets with built-in AI Agents providing engineers with real-time solutions that help them avert the risk of unscheduled stoppages while dramatically increasing process tool uptime, extending asset life, and greatly reducing wafer scrap.”
About the Manufacturing AI Applications Framework
The PODO AI framework includes a broad range of applications, including Asset Performance Management (APM), Asset Benchmarking Performance (ABP), Process Analytics, and Energy Management System (EMS) applications, all optimized to deliver unmatched detection, prediction, and prescription capabilities across the production ecosystem, real-time, 24/7.
To ensure an “always on” semiconductor wafer production environment, fab managers need a true picture of their entire fab operation. Providing a holistic view of the health and performance of critical sub-fab, cleanroom, and facility equipment and processes has been elusive until now. Hundreds of millions of dollars are lost each year in unplanned downtime, disrupting production processes, reducing productivity and yields, which depresses overall profitability. With Aidentyx’s manufacturing AI suite, semiconductor manufacturers and their suppliers substantially reduce the risk of unplanned downtime by identifying potential failures in the sub-fab, and/or cleanroom before they occur. This reduces the heavy costs of last-minute parts and maintenance and, most importantly, alleviates potential production losses.
Empowering Engineers by Connecting Customers to All Assets Regardless of Vendor or Type
PODO AI is a vendor-agnostic AI software and sensors system that allows data collection, detection, and prediction of fab equipment, sub-fab systems, and production processes. The system empowers sub-fab and cleanroom engineering teams to optimize maintenance strategies, to automatically perform data collection, fault detection, and perform predictive analytics of equipment and processes in real-time. In addition, new root cause analysis and knowledge-based applications powered by GenAI technology will provide timely prescriptive-based solutions to problems across the fab. “We can manage and monitor any number of assets to provide engineers with a status of the health of all their assets, regardless of vendor,” said Jason Kim, CEO, Aidentyx. “Together, the system provides early warning detection of potential failures, and applications expertise for proactively addressing these issues, particularly those issues that impact process tool uptime.”
“There’s no doubt that sub-fab failures greatly impact process tool performance in the cleanroom and we can see through our research working with customers and partners that we can coordinate seemingly disparate manufacturing areas to eliminate the risk of costly and unpredictable stoppages to maximize profitability,”” commented Stewart Chalmers, Chief Business Officer of Aidentyx. “This is finally a great opportunity for fab managers and their suppliers to see the full picture and to have digital tools that empower them to act proactively.”
Fab of the Future
The adoption of new AI-powered systems and solutions is accelerating, which helps to support the creation of an AI-ready workforce. This is key to ensuring the long-term competitiveness of fabs everywhere. According to W.K. Choi, Chairman and CEO of Aidentyx’s strategic partner, BISTelligence, “Any semiconductor tool maker, or sub-system supplier that doesn’t embrace AI analytics in their products and service offerings will fail. OEMs will need to quickly embrace digital transformation in their offerings to their customers or risk being left behind.”
AI Applications Framework Overview
The new AI Framework provides a common data platform for engineering analytics applications of all types (now and in the future), which are orchestrated by virtual AI Agents to exceed quality and production goals. The system’s AI-powered analytic applications seamlessly integrate with any external enterprise system. Capabilities include:
- Real-Time Asset Health Index Around the Clock
- Asset Performance Details
- Asset Performance vs. Fleet Performance Analysis
- AI Data Analysis
- Performance Anomaly Detection
- Anomaly Root Cause Analysis
- Failure Classification
- Energy Management and Optimization
- Asset Longevity Prediction
- Asset Health Maintenance Operation Management
- Workflow Automation (Closed Loop)
- Prescriptive Analytics
- Process Performance Management
- Speedy Data Collection and Integration
- Global Visibility of Cleanroom, Sub Fab, Facility, Production Line, and Processes
- Real-time Alerts and Notifications of Issues
- Customizable Alarming Options to Minimize False Alarms
- Real-Time Detection Model Optimization