ChinaManufacturing2026 – ongoingSuzhou Universal
AI agents for quality inquiry and production scheduling
Suzhou Universal cut scheduling errors by 70% with AI agents linked to its factory systems
Problem
Production, planning and quality data sat in separate MES, ERP, APS and quality systems. Staff spent up to 30 minutes on a cross-system query, and defect analysis and scheduling changes were largely manual.
Approach
Suzhou Universal built an agent platform on IBM watsonx Orchestrate with the partner X-POWER (Suzhou) Information Technology. It joined the factory systems into one data layer. Two agents were deployed. A quality agent monitors quality, analyzes defects and finds root causes. A scheduling agent lets planners query and change production plans in natural language, and changes are checked in a draft environment before they run.
Results
Takeaway
Connecting the plant's separate systems into one data layer came first. The agents then answered questions that had needed manual work in each system.
Categories
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- AI agents
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
IBM describes an agentic AI platform at the precision component maker Suzhou Universal, built on watsonx Orchestrate. It reports faster queries and analysis, better root cause accuracy, over 30% better scheduling efficiency and 70% fewer scheduling errors.
Full details are in the original sources.
