TaiwanEnergy and utilities2026 – ongoingCPC Corporation, Taiwan
Dalin Refinery AI Assistant
CPC's Dalin refinery reported 90% less time interpreting operating data with a local AI assistant
Problem
Refinery staff manually compared laboratory and operating data held in separate systems. Sensitive production information could not be sent to external cloud AI services.
Approach
Dalin developed an assistant that extracted data visible in browser-based systems and passed it to local language models inside its VPN. The assistant explained anomalies and sent readable messages to LINE groups. It also analyzed screenshots to help staff examine operating problems.
Results
Takeaway
The refinery acknowledged that missing reference documents could lead to incomplete operating suggestions.
Categories
- Size
- Large enterprise
- Industry
- Energy and utilities
- Business function
- Production and qualityInternal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AIComputer vision
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The refinery team's account described an internally developed assistant using browser data and local models. It reported 90% shorter interpretation time and anomaly matrices produced in 5 seconds, while acknowledging hallucination risks.
Full details are in the original sources.
