ERP (L4, days-months): business planning, finance, orders, customers — SAP, Oracle. MES (L3, minutes-hours): operations execution, recipes, traceability, OEE — Siemens Opcenter, Aveva, Werum. Historian (L2, milliseconds-seconds): time-series sensor data archive — OSIsoft PI, Aveva PI, AspenTech IP.21. Three distinct layers with B2MML integration. OEE specialists (TeepTrak Pulse) at L3 focused on measurement.
Three categories of industrial software are often confused: ERP, MES, and Historian. Each operates at a different ISA-95 level, handles different data types, and serves different business questions. ERP (Enterprise Resource Planning) addresses “what should we produce and when, and how does it impact finance?”, MES (Manufacturing Execution System) addresses “how are we executing production right now, and is it according to the recipe?”, Historian addresses “what was the exact temperature/pressure/flow at 14:32:17.428 yesterday?”. Understanding these distinctions is critical for architectural decisions, vendor selection, and integration strategy. This guide details each layer, their typical vendors, integration patterns, and the positioning of OEE specialists like TeepTrak Pulse in this landscape.
ERP: Enterprise Resource Planning (ISA-95 L4)
ERP covers business planning and logistics at enterprise level. Time horizon: days to months. Data: aggregate orders, inventory balances, finance entries, customer records, supplier records, BOMs, production schedules at master level.
| Major ERP vendors 2026 | Product | Market |
|---|---|---|
| SAP | S/4HANA (cloud or on-premise) | Global leader, large enterprise |
| Oracle | Oracle Cloud ERP (Fusion) | Large enterprise, financial services |
| Microsoft | Dynamics 365 Finance + Supply Chain | Mid-market, cloud-first |
| Infor | Infor CloudSuite Industrial / LN | Mid-market, industry-specific |
| Epicor | Epicor Kinetic | Mid-market manufacturing |
| IFS | IFS Cloud (ERP + EAM + Service) | Asset-intensive industries |
| Sage | Sage X3 / Intacct | SMB to mid-market |
| Workday | Workday Adaptive Planning | HR-centric, finance |
| Odoo | Odoo Enterprise/Community | SMB, open-source |
| NetSuite (Oracle) | Oracle NetSuite | Cloud SMB |
Typical ERP functional scope: Finance & Accounting (GL, AP, AR), Sales & Distribution, Procurement, Production Planning (MRP), Master Data Management, Human Resources, Asset Management (EAM). Time horizon: typically daily/weekly aggregations from MES, monthly closes for finance.
MES: Manufacturing Execution System (ISA-95 L3)
MES covers operations execution at site level. Time horizon: minutes to hours. Data: unitary work orders, recipes per batch, equipment states, operator actions, in-process quality checks, traceability genealogy, OEE in real-time.
| Major MES vendors 2026 | Product | Industry focus |
|---|---|---|
| Siemens | Opcenter Execution Discrete/Process/Pharma/Semiconductor | All industries |
| Aveva (Schneider Electric) | Aveva MES (ex-Wonderware) | Process (chemicals, F&B, mining) |
| Rockwell Automation | FactoryTalk ProductionCentre + PharmaSuite | Discrete, pharma |
| Werum (Körber) | PAS-X MES | Pharma, biotech |
| Dassault Systèmes | Apriso (3DEXPERIENCE) | Aerospace, automotive |
| GE Vernova | Proficy Smart Factory | Process, energy |
| Honeywell | Honeywell Forge MES | Process, refining |
| SAP | Digital Manufacturing (DM) | SAP customers |
| Plex (Rockwell) | Plex Smart Manufacturing Platform | Discrete cloud SaaS |
| iBASEt | Solumina | Aerospace, defense |
| Critical Manufacturing | cmNavigo | Semiconductor, electronics |
| TeepTrak (OEE specialist) | Pulse | OEE focus, plug-and-play |
Typical MES functional scope: Recipe Management (ISA-88 for batch), Order Execution, Resource Allocation, Equipment Performance / OEE, Quality / SPC, Maintenance (basic), Traceability / Genealogy, Inventory at work-in-progress level. Time horizon: real-time to shift-end.
Historian: Process Time-Series Database (ISA-95 L2)
Historian covers high-frequency process data archival from sensors and controllers. Time horizon: milliseconds to seconds (with long-term storage). Data: temperature, pressure, flow, level, vibration, voltage, current, RPM — all numeric time-series, sampled 1-1000 Hz typical.
| Major Historian vendors 2026 | Product | Strengths |
|---|---|---|
| Aveva (was OSIsoft) | PI System (PI Server, AF, Vision, Integrator) | De facto industrial standard, 8000+ sites worldwide |
| AspenTech (Emerson) | IP.21 InfoPlus.21 | Refining, petrochemicals, chemicals |
| GE Vernova | Proficy Historian | Power generation, GE installed base |
| Honeywell | Uniformance PHD | Honeywell Experion DCS ecosystem |
| Siemens | SIMATIC Process Historian | Siemens PCS 7 ecosystem |
| Rockwell Automation | FactoryTalk Historian (powered by OSIsoft) | Rockwell ecosystem |
| Yokogawa | Exaquantum | CENTUM VP DCS ecosystem |
| Canary Labs | Canary Historian | Cost-effective alternative |
| InfluxData | InfluxDB | Open-source / cloud time-series |
| TimescaleDB | TimescaleDB (PostgreSQL) | Open-source, SQL-compatible |
| AWS | Timestream | Cloud-native AWS |
| Microsoft | Azure Data Explorer (ADX) / TSI | Cloud-native Azure |
Typical Historian functional scope: raw tag collection at high frequency, compression (swinging door, dead-band), long-term archival (years), interpolation/aggregation queries, trend visualization, alarms, integration with BI tools. Modern historians extend to “operational data lake” with edge collection + cloud storage.
Download the white paper
Enter your email address to receive our White Paper
Decision matrix: which system handles which question?
| Business question | System | Time horizon | Example query |
|---|---|---|---|
| What is our financial result this quarter? | ERP | Months | SELECT revenue FROM finance WHERE quarter=’Q3-2026′ |
| What is our inventory balance for SKU X? | ERP | Daily | Get inventory for material X |
| What customer order are we currently fulfilling? | ERP→MES | Daily | Match work order to customer order |
| What is the current OEE of line 5? | MES / OEE specialist | Real-time | Current shift OEE = A × P × Q |
| What recipe is running on reactor R-302? | MES | Real-time | Active batch record for R-302 |
| Why did batch B-456 fail quality? | MES + Historian | Last week | Genealogy + process trends during batch |
| What was the temperature at 14:32:17.428 yesterday? | Historian | Milliseconds | SELECT value FROM tags WHERE tagname=’TI-302′ AND timestamp=’2026-12-15 14:32:17.428′ |
| What is the trend of vibration on pump P-101 last month? | Historian | Hours/days | Aggregated trend of VIB-P101 |
| Did we ship on time vs production plan? | ERP→MES | Daily | Compare planned vs actual production |
| Who operated machine M-205 during shift 2 yesterday? | MES | Yesterday | Operator login records |
Integration patterns: ERP ↔ MES ↔ Historian ↔ OEE
Modern manufacturing architecture follows ISA-95 hierarchy with B2MML XML messages (or modern REST/JSON adaptations):
- ERP → MES (downward): ProductionSchedule (production orders, master recipes, BOMs), ProductDefinition (specifications), MaterialDefinition (raw materials, packaging)
- MES → ERP (upward): ProductionPerformance (quantities produced, scrap, yield), MaterialConsumption (actual BOM consumption), ProductionResponse (work order completion)
- SCADA/PLC → Historian (upward): high-frequency tag values via OPC UA, MQTT, native protocols (Modbus, Ethernet/IP, PROFINET)
- Historian → MES (lateral): aggregated process data for batch records, OEE calculations, quality checks (data points sampled at relevant intervals)
- MES → Historian (lateral): event-based: order start/end, quality results, deviations, OEE events
- OEE specialist (TeepTrak Pulse) integration: read-only from SCADA/PLC via OPC UA (cycle data) + manual inputs (operator categorization), write to ERP/MES summary OEE KPIs via REST API
Architectural patterns: replacement vs coexistence
Pattern A: Coexistence (most common, 70% of brownfield deployments)
Keep existing ERP + existing Historian, add or upgrade MES. Pattern: SAP S/4HANA + Aveva MES + OSIsoft PI System (process industries) OR SAP S/4HANA + Siemens Opcenter Execution Discrete + Proficy Historian (discrete). OEE specialist (TeepTrak Pulse) layered for real-time visualization where MES is too slow or absent.
Pattern B: ERP-centric (SAP Digital Manufacturing approach)
SAP S/4HANA + SAP Digital Manufacturing (MES) + SAP-integrated Historian (third-party OSIsoft PI). Reduces integration complexity (single vendor), but locks into SAP ecosystem and may not match operational depth of best-of-breed MES.
Pattern C: Cloud-native (modern greenfield)
Cloud ERP (NetSuite, Microsoft Dynamics 365) + Cloud MES (Plex, Tulip, Litmus Edge, MachineMetrics) + Cloud Historian (AWS Timestream, Azure ADX, InfluxData Cloud). Faster deployment (3-6 months vs 18-24 months), lower initial cost, less customization. Best for SMB and new sites.
Pattern D: Hybrid with OEE specialist (TeepTrak pattern)
Existing ERP + Existing MES (or no MES) + Existing Historian + Add TeepTrak Pulse for OEE on critical equipment. Deployment 8-12 weeks for OEE layer regardless of underlying MES/Historian complexity. Hutchinson 40-site case: this pattern, with TeepTrak Pulse adding real-time OEE visibility across heterogeneous MES landscape.
FAQ: MES vs ERP vs Historian
Can ERP replace MES?
Generally no. ERP (SAP, Oracle, Microsoft Dynamics) handles business planning at daily/monthly horizons, while MES handles execution at minute/hour horizons with deep operational features (recipes, traceability, OEE). SAP Digital Manufacturing attempts to extend ERP into MES territory but typically lacks depth for complex manufacturing (pharma batch records, semiconductor wafer tracking, aerospace as-built). Best practice: distinct ERP + MES layers with B2MML integration.
Can Historian replace MES?
No. Historian stores time-series sensor data; MES executes work orders with workflow logic, recipes, traceability. They are complementary: MES reads aggregated/sampled values from Historian to populate batch records or OEE calculations. OSIsoft PI is the dominant Historian; it does not provide MES workflow capabilities.
What is the difference between MES and SCADA?
SCADA (ISA-95 L2): real-time process supervision and control, alarms, HMI for operators. Time horizon: seconds. MES (ISA-95 L3): operations management, work orders, recipes, OEE, traceability. Time horizon: minutes-hours. Many vendors offer integrated SCADA+MES (Aveva InTouch + Aveva MES = unified suite), but conceptually distinct layers.
How do MES and Historian integrate?
Historian collects high-frequency tags from PLCs/SCADA (1-100 Hz typical). MES queries Historian for batch record context: average temperature during batch step 3 = 76.5°C from 14:30 to 14:45. MES events (batch start, batch end, deviations) typically also written to Historian for unified data lake. Modern integration: REST API queries from MES to Historian, or shared data platform (e.g., Aveva Data Hub for unified MES + PI access).
What is the role of OEE specialist (TeepTrak Pulse) in this landscape?
OEE specialists are L3 specialists focused on real-time OEE measurement (A × P × Q) with operator categorization (Six Big Losses). They complement full MES by providing fast deployment (8-12 weeks) and standardized OEE measurement across heterogeneous MES landscape (multi-site groups). TeepTrak Pulse integrates read-only with existing MES/SCADA via OPC UA and writes aggregated KPIs to ERP via REST API.
What is the typical integration cost ERP-MES?
€200-800k for B2MML integration between SAP S/4HANA and major MES (Siemens, Aveva, Werum). Includes: schema mapping, transaction definition (production schedule + performance + master data), middleware (SAP PI/PO or MuleSoft or Boomi), testing, hypercare. Multi-site deployments amortize this cost over multiple sites.
Can I run a manufacturing site with only ERP and Historian (no MES)?
Possible for low-complexity sites (single product, manual operations, no traceability requirements). For most modern industrial sites: MES adds critical value (recipes for process, OEE for discrete, traceability for regulated). Without MES, operators rely on spreadsheets, leading to weak data quality and compliance gaps.
What about Data Lake / Data Warehouse vs Historian?
Modern data lake (Snowflake, Databricks, AWS Lake Formation, Microsoft Fabric) aggregates data from ERP + MES + Historian + other sources for analytics. Historian remains the system of record for high-frequency sensor data; data lake consumes downsampled/aggregated data for cross-functional analytics. Coexistence pattern: Historian for OT, Data Lake for analytics.
Which integration standards apply between layers?
ERP ↔ MES: B2MML (XML, ISA-95 based), or REST/JSON adaptations, increasingly OPC UA Companion Specifications. SCADA/PLC ↔ Historian: OPC UA standard, native PLC protocols (Modbus TCP, Ethernet/IP, PROFINET), MQTT for IIoT. MES ↔ Historian: typically vendor-specific REST APIs or OPC UA. All major MES support B2MML for ERP integration.
How does Industry 4.0 / IIoT change this architecture?
Industry 4.0 / IIoT adds: edge computing layer (between L2 and L3), digital twins (overlay on equipment), asynchronous publish/subscribe (MQTT, Kafka), AI/ML services consuming historian + MES data. The ISA-95 hierarchy remains but with more horizontal data flows (peer-to-peer between equipment) and edge-cloud-hybrid deployments. Cloud-native MES (Plex, Tulip, Litmus) and cloud Historian (AWS Timestream) gain ground.
Conclusion
ERP, MES, and Historian occupy distinct layers of the ISA-95 hierarchy with complementary roles: ERP for business planning (L4, days-months), MES for operations execution (L3, minutes-hours), Historian for process time-series data (L2, milliseconds-seconds). Modern manufacturing architectures combine all three with B2MML/REST integration. OEE specialists like TeepTrak Pulse position at L3 with specific focus on OEE measurement, complementing rather than replacing full MES. Coexistence pattern (70% of deployments) typical for brownfield; cloud-native pattern emerging for greenfield. Best practice: distinct layers, vendor-neutral integration standards (B2MML, OPC UA).
Next step: download the TeepTrak MES vs ERP vs Historian comparison whitepaper or request a free architecture maturity assessment.
0 Comments