MES as the Execution Layer Between ERP and SCADA

Manufacturing software is usually discussed in layers.

At the top, ERP systems manage business-level planning: orders, materials, purchasing, inventory, finance, and delivery commitments.

At the bottom, SCADA and control systems manage the real-time industrial process: machines, sensors, PLCs, alarms, HMI screens, and process variables.

Between these layers sits a difficult problem: production execution.

This is where work orders become operator tasks, materials become finished goods, downtime becomes performance loss, and quality checks become part of the product history.

That middle layer is usually handled by MES, or Manufacturing Execution System.

Why ERP and SCADA are not enough on their own

ERP and SCADA solve different problems.

ERP is designed around business transactions. It knows what should be produced, what materials are available, what orders exist, and what the business expects from the factory.

SCADA is designed around process supervision. It knows what is happening in machines, lines, utilities, and industrial equipment right now.

The gap appears when the factory needs to answer execution-level questions:

• Which work order is currently running on this line?

• Which operator performed this step?

• Which batch or lot was consumed?

• Why did the line stop?

• Was the quality check completed?

• What was the actual cycle time?

• Which finished goods are affected by a material issue?

• Did production follow the approved routing or recipe?

ERP usually does not capture this level of shop-floor detail in real time.

SCADA usually does not understand business objects such as production orders, routes, batches, quality records, or material genealogy.

MES connects these worlds.

A simple MES architecture

A simplified MES architecture can be represented like this:

+-------------------------------+
|              ERP              |
| Orders, inventory, planning   |
+---------------+---------------+
                |
                v
+-------------------------------+
|              MES              |
| Scheduling, execution, OEE,   |
| quality, traceability, reports|
+---------------+---------------+
                |
                v
+-------------------------------+
|          SCADA / HMI           |
| Machines, alarms, tags,       |
| process visualization         |
+---------------+---------------+
                |
                v
+-------------------------------+
|        PLCs / Sensors / Edge   |
| Real-time industrial signals  |
+-------------------------------+

The MES layer does not replace ERP or SCADA.

It coordinates execution between them.

ERP sends production requirements. SCADA and automation systems provide real-time production data. MES connects both sides and maintains the operational record of what actually happened.

Core functions of an MES layer

A practical MES implementation usually includes several functional areas.

1. Production order execution

ERP may create or approve the production order, but MES manages how that order is executed on the shop floor.

This includes:

• Dispatching work to lines, cells, or stations

• Assigning operations to shifts or operators

• Tracking order progress

• Recording start and stop events

• Capturing actual output

• Handling exceptions during execution

The important difference is that MES deals with actual production activity, not only planned production activity.

2. Routing, recipes, and BOM control

Manufacturing execution depends on more than a work order number.

The system needs to know which operations must happen, in what sequence, using which equipment, under which parameters, and with which materials.

For discrete manufacturing, this may involve routings, stations, components, and serial numbers.

For batch or process manufacturing, this may involve recipes, formulas, process parameters, tanks, lines, and batch records.

This is where MES becomes the structured execution model for production.

3. Scheduling and capacity constraints

A production plan is only useful if it reflects real constraints.

Machines have capacity limits. Operators work shifts. Materials may not be available. Maintenance windows reduce available time. Changeovers and setup time affect throughput.

MES scheduling should account for constraints such as:

• Equipment availability

• Labor availability

• Setup and changeover time

• Material readiness

• Production priorities

• Maintenance windows

• Line-specific capabilities

This makes scheduling closer to the real factory instead of a theoretical production calendar.

4. OEE and downtime analysis

OEE is commonly used to measure production performance through three components:

• Availability

• Performance

• Quality

MES can calculate OEE by combining production context with real-time machine data.

For example, a line stop from SCADA becomes more useful when MES can connect it to:

• The active work order

• The product being made

• The shift

• The operator group

• The downtime reason

• The affected quantity

• The related maintenance event

Without this context, downtime is just a signal.

With MES context, downtime becomes an operational event that can be analyzed and improved.

5. Quality management during production

Quality should not be disconnected from execution.

If inspections are recorded only after production, quality data becomes delayed and incomplete.

MES can bring quality checks into the production workflow through:

• In-process inspections

• SPC data collection

• Defect tracking

• Operator confirmations

• Deviation records

• Electronic batch records

• Audit-ready production history

This allows quality events to be connected to specific materials, equipment, operators, process values, and production steps.

6. Traceability and genealogy

Traceability is one of the strongest reasons to implement MES.

A complete traceability model answers two directions of questions.

Backward traceability:

Finished product -> batch -> operations -> equipment -> materials -> suppliers

Forward traceability:

Raw material lot -> consumed batches -> finished products -> shipments

This is critical when dealing with recalls, audits, defects, warranty claims, regulated production, or customer complaints.

The goal is not just to store data. The goal is to reconstruct the production history quickly and reliably.

7. Reporting and production records

MES reporting is different from general business reporting.

It needs to combine production context with real-time and historical data.

Common reports include:

• Shift reports

• Production summaries

• OEE reports

• Downtime reports

• Batch reports

• Quality reports

• Material consumption reports

• Traceability reports

The value of these reports depends heavily on the quality of the underlying data model.

If production events, materials, equipment, and work orders are not connected consistently, reporting becomes another manual process.

Why ISA-95 is useful

MES projects often become difficult when each plant models production differently.

One site may use different names for areas, lines, work centers, operations, equipment, and materials. Another site may have a different structure entirely.

This creates problems for ERP integration, reporting, analytics, and multi-site rollout.

ISA-95 provides a standard way to think about enterprise-control integration and manufacturing operations.

It helps structure concepts such as:

• Enterprise

• Site

• Area

• Production line

• Work cell

• Equipment

• Material

• Personnel

• Operations

• Production segments

The goal is not to force every factory into the same process.

The goal is to create a consistent model that software systems can understand.

A consistent MES data model makes it easier to integrate with ERP, compare performance across sites, reuse templates, and build long-term analytics.

Deployment patterns: cloud, on-premise, edge, and hybrid

MES deployment depends on the plant environment.

There is no single model that works for every manufacturer.

Cloud MES

Cloud deployment can be useful for multi-site access, centralized reporting, and easier maintenance.

It is often attractive when factories need shared visibility across locations.

On-premise MES

On-premise deployment is still common in manufacturing.

Reasons include data control, security policy, latency, regulatory requirements, and limited external connectivity.

Edge MES

Some execution logic may need to run close to production equipment.

Edge deployment can help when the factory needs local operation even if the wider network connection is unstable.

Hybrid MES

Hybrid deployment combines local execution with centralized visibility.

In this model, the plant can continue running locally while central systems receive synchronized data when connectivity is available.

MES should reduce integration complexity

Many factories end up with fragmented industrial software stacks.

One system handles SCADA. Another handles MES. Another handles maintenance. Another handles analytics. Another handles reporting. Another handles edge data collection.

Over time, this can create:

• Duplicated tag databases

• Repeated integrations

• Inconsistent equipment naming

• Separate user interfaces

• Unclear ownership of data

• Expensive middleware

• Slow reporting cycles

A cleaner architecture connects MES, SCADA, reporting, maintenance, and analytics around a shared operational model.

For example, machine data from SCADA can feed OEE calculations. Downtime can create maintenance workflows. Quality deviations can be linked to process parameters. Traceability reports can combine material, equipment, and operator data.

This is where MES becomes more than a production database.

It becomes the execution backbone of the factory.

A technical checklist for MES evaluation

When evaluating MES architecture, it is useful to ask technical questions before discussing features.

For example:

• Does the system support production orders, routes, recipes, BOMs, and work instructions?

• Can it connect real-time machine data with production context?

• Does it support OEE calculation with downtime reason handling?

• Can it maintain full batch, lot, or serial genealogy?

• Does it support in-process quality checks and electronic records?

• Is the data model consistent with ISA-95 concepts?

• Can it integrate with ERP without heavy custom middleware?

• Can it run in cloud, on-premise, edge, or hybrid environments?

• Can implementation teams modify workflows and reports without rebuilding the system?

• Can the same model be reused across multiple sites?

A useful MES is not only a set of screens.

It is a structured execution model for production.

For a concrete example of how common MES modules are grouped in an industrial platform, this MES software architecture overview shows scheduling, OEE, traceability, quality, reporting, HMI, and deployment options in one system.

Final thoughts

MES exists because manufacturing needs more than business planning and machine supervision.

ERP defines what should happen.

SCADA shows what is happening.

MES records, coordinates, and validates what actually happened during production.

The technical challenge is to connect these layers without creating another isolated system.

A good MES architecture should make production execution visible, traceable, measurable, and adaptable while keeping the data model clean enough for integration, reporting, and long-term improvement.