TAIPEI (Taiwan News) — ASE Technology Holding Co. is increasing the influence of AI in its factories through academic partnerships, turning research into tools for its advanced semiconductor packaging and testing lines.
The company has worked with six universities: National Central University, National Cheng Kung University, National Sun Yat-sen University, National Yang Ming Chiao Tung University, National Tsing Hua University, and National Taiwan University of Science and Technology. Twelve projects cover applications such as generative AI, digital twins, predictive maintenance, and cybersecurity, according to CTEE.
The work is aimed at improving production efficiency and process quality while making operations more resilient to disruptions and climate-related risks. ASE said the broader goal is to shift from addressing problems after they occur to preventing them in advance on its chip packaging lines.
Some of the research has already been turned into AI assistants for factory workers. One tool gathers video, image, and text content from different platforms and automatically produces daily summaries, making it easier to find and review information.
Another helps workers troubleshoot equipment by drawing on a dedicated knowledge base. The assistant is designed to speed up fault identification and reduce repair time.
A third tool checks program code and automatically corrects parts that do not meet required standards. ASE said the rate of code returned for revisions has fallen below 10% since the tool was introduced.
The technology is also changing how ASE manages production. Smart scheduling tools can simulate changes in capacity as orders shift, while real-time equipment monitoring can identify abnormalities within milliseconds and allow workers to make adjustments before they lead to unexpected downtime.
Digital twin technology gives ASE a way to test operations in a virtual environment. The company is using it for production scheduling and automated guided vehicles, helping shorten the time needed to plan and test changes.
AI is also being applied beyond production. AI-powered cameras can issue warnings about potential hazards and track people and vehicles, while AI agents can automate cybersecurity checks for industrial control systems and support recovery when problems are detected.
It is also being used to assess the physical environment of ASE's facilities. The company is analyzing climate-related risks and vulnerabilities at its sites, while airflow simulations in cleanrooms are being used to identify potential ways to reduce energy consumption.
Virtual modeling is another way ASE is working to anticipate problems. The company is combining a 3D simulation platform with AI to build digital models of its facilities. The systems can help detect equipment problems before they occur, monitor product yields in real time, and automate quality inspections.
