In modern manufacturing, simply collecting data is not enough. The real value comes from storing the right data in the right place so it can serve its intended purpose—whether that’s real-time visibility, long-term machine performance analysis, or enterprise-level business insights.
This article explains the roles of the Unified Namespace (UNS), the Historian, and the Lakehouse, and why each storage type is essential in a smart manufacturing data strategy.
UNS: The Real-Time Nervous System
The Unified Namespace (UNS) acts as the backbone for real-time data orchestration. It provides a single source of truth for the current state of operations, ensuring that both operators and systems see the same picture of what’s happening on the shop floor.
What belongs in UNS:
- Unfinished transactions (e.g., production orders currently running, jobs in progress)
- Live events (e.g., machine running/idle state, downtime start, operator login)
- Recently finished transactions (e.g., a work order just completed, a changeover just finished)
- Contextual snapshots (e.g., inbound/outbound queue, current BOM revision, who is on shift)
Purpose:
- Provide real-time visibility for operators, supervisors, and connected systems
- Act as a communication layer for orchestration between OT and IT
- It does not serve as long-term storage but rather functions as the real-time nervous system of manufacturing
Historian: The Machine Memory
A Historian is designed specifically for high-frequency OT data. It excels at capturing time-series signals with high precision, efficient compression, and robust retention strategies.
What belongs in a Historian:
- Production parameters (temperature, pressure, speed, setpoints)
- Machine health signals (vibration, motor current, bearing temperature)
- Energy usage (kW, kWh, power factor)
- Good/reject counters, cycle counts
- Downtime events (start/end, reason codes)
- Utility/environmental parameters (air, water, steam, compressed air)
Purpose:
- Enable trend analysis, SPC, and golden batch studies
- Provide data for root cause investigations and predictive maintenance
- Long-term retention of 3–7 years at high resolution
- Functions as the memory of machines and operations
Lakehouse: The Enterprise Brain
Lakehouse รวมข้อดีของ Data Lake และ Data Warehouse เข้าด้วยกัน เป็นศูนย์กลางของข้อมูลสำหรับ analytics, compliance, และ AI/ML เชื่อมทั้งข้อมูล OT และ IT เข้าด้วยกัน
ข้อมูลที่ควรอยู่ใน Lakehouse
- A Lakehouse combines the flexibility of a data lake with the governance of a data warehouse. It is the central hub for analytics, compliance, and AI/ML, bringing together both OT and IT data.
- What belongs in a Lakehouse:
- Production events (order release, yield, scrap, start/end of operations)
- WIP and inventory movements (goods issue to production, goods receipt, shipments)
- Quality data (SPC measurements, NCRs, CAPAs, deviations)
- Maintenance records (work orders, spare part transactions)
- Energy cost data (tariff tables, cost centers)
- Pure ERP/IT data (sales orders, invoices, purchase orders, AP/AR, GL, payroll, employee master)
- Purpose:
- Provide a foundation for business analytics and forecasting
- Support compliance reporting and regulatory audits
- Serve as a data hub for machine learning and AI use cases (predictive quality, demand forecasting)
- Store data for the long term (7–10+ years)
- Functions as the enterprise brain, aligning operational and business perspectives
- Summary
- UNS = Real-time nervous system → unfinished transactions, live events, snapshots
- Historian = Machine memory → high-frequency OT signals, long-term trends, SPC
- Lakehouse = Enterprise brain → analytics, compliance, AI/ML, business transactions
