TaiwanManufacturing2026 – ongoingASE Technology Holding
STAR multimodal AI agent for equipment maintenance diagnosis
ASE's STAR maintenance agent cut fault diagnosis from over an hour to minutes in the units that use it
Problem
Engineers diagnosing faults in semiconductor packaging and test equipment often needed more than an hour to find the cause of an abnormality.
Approach
ASE built STAR, a maintenance guidance system based on a multimodal generative AI agent with retrieval over knowledge sources. It reads equipment operating data, process parameters and fault codes, analyses the cause and gives repair steps. It started in one of five business units at the Kaohsiung plant, and two more units have since joined.
Results
Takeaway
ASE reported good results in the first unit before widening use. The aim is all five Kaohsiung units within a year, which is a target, not a result.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Generative AIAI agents
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Describes ASE's STAR system and its move from a passive query tool to an agent. Diagnosis time fell from over an hour to the minute level. Three of five Kaohsiung units use it, with all five targeted within a year.
Full details are in the original sources.