Standardizing Industrial Data for SCADA, AI & Ignition Edge

Modern plants run on data - but when every machine, vendor, and integrator names tags differently, that data becomes fragmented. This lack of consistency makes it difficult to deploy OEE, alarms, AI analytics, or reusable templates across sites.

Tag Harmonization solves this problem by creating a unified, consistent, and intelligent naming structure for industrial systems.

Why Tag Harmonization Matters

Most SCADA systems operate with chaotic tag structures:

• Different names for the same measurement
• No consistent units or descriptions
• Vendor-specific folders and registers
• Templates that break from site to site
• Complex UDTs that cannot be reused

This makes scaling digital transformation nearly impossible.

Harmonized tags enable:

• Reusable User Defined Templates (UDTs) for Ignition Edge
• Standardized alarms and KPIs across all equipment
• AI-ready data for anomaly detection & predictive analytics
• Fast onboarding of new lines, machines, or facilities
• Consistent reporting and benchmarking at fleet scale

How We Harmonize Tags

Our approach combines semantic embeddings, vector databases, and functional equivalence modeling to automatically detect which tags "mean the same thing," even if they look completely different.

1. Semantic Embeddings

Each tag name and description is converted into a high-dimensional vector representing its meaning.

2. Cosine Similarity Matching

We calculate how similar two tags are by comparing their vectors.

Example:

• Voltage_PhaseA ↔ MMXU1.PhV.phsA.instMag.f → 0.91 similarity

3. Functional Equivalence Rules

We categorize signals (voltage, current, temp, status) based on units, PLC register type, vendor conventions, and logical domain.

4. Automated Template Generation

Once tags are harmonized, we generate:

• Ignition Edge UDTs
• Standard folders
• Derived tags
• Alarm blocks
• KPI models (OEE, availability, thermal health, etc.)

This turns messy tag lists into repeatable, scalable SCADA templates.

Results for Customers

• 10× faster Ignition Edge deployments
• 60-80% reduction in engineering hours
• Fleet-level standardized SCADA models
• AI models trained on consistent inputs
• Drop-in UDTs usable across any vendor

Sample Mapping Examples

Voltage Measurement Mapping

Temperature Sensors

Status / Boolean Signals

Harmonized Naming Structure

A recommended structure used in successful deployments:

{AssetType}/{AssetId}/{Domain}/{Signal}/{Attribute}

Examples:

• INVERTER/1101/VOLTAGE/PhaseA/Instantaneous
• INVERTER/1101/TEMPERATURE/Module01/Average
• INVERTER/1101/STATUS/Run/State
• TRANSFORMER/01/POWER/Active/Value

This structure works flawlessly with UDTs and can be reused across SMA, Sungrow, TMEIC, CATL, GE, Yaskawa, Siemens, Mitsubishi - without rebuilding templates.

Multi-Vendor Fleet Normalization

Before harmonization:

• INV1_MMxU1.PhV.phsA.f
• INV_A1_VoltA
• INVA01.PV.A
• 1A-LineVoltageA

After harmonization:

• Voltage.PhaseA

One meaning. One structure. Infinite scalability.