How can I use manufacturing execution systems to integrate with ERP and equipment on the plant floor?

Manufacturing Execution Systems (MES) sit between the ERP layer and the plant floor, acting as the coordination and data-collection layer that turns production plans into executed work and real-time visibility.

Typical integration pattern

1. ERP sends business-level demand to MES

ERP usually owns:

• customer orders
• production orders / work orders
• material requirements
• BOMs and routings
• inventory and purchasing

MES receives from ERP:

• what to make
• how much to make
• due dates / priorities
• product specs and routing definitions
• materials and lot/serial requirements

This is often done through APIs, web services, message queues, or file-based interfaces, depending on the systems involved.

2. MES orchestrates execution on the shop floor

MES translates ERP orders into shop-floor actions:

• dispatch work to lines, cells, or machines
• manage operator instructions and work steps
• enforce electronic work instructions
• track WIP, genealogy, and traceability
• collect labor, quality, and downtime data
• manage deviations and approvals

MES becomes the system of record for production execution.

3. MES connects to equipment and controls systems

MES typically integrates downward with:

• PLCs
• SCADA/HMI systems
• historians
• machine sensors and IIoT gateways
• robotics or automated material handling systems

This layer is used to:

• read machine states, counts, alarms, and cycle times
• send setpoints or recipe parameters
• trigger job start/stop or changeover events
• capture process data automatically
• validate execution against the plan

Common industrial protocols include:

• OPC UA
• MQTT
• Modbus
• Ethernet/IP
• Profinet
• vendor-specific APIs

Recommended architecture

A common design is:

ERP → MES → SCADA/PLC/Equipment

With MES as the integration hub:

• upward: receives demand and reports production status back to ERP
• downward: sends work instructions and receives live machine data

In more advanced setups, you may also use an integration layer or middleware between systems to:

• normalize data
• handle retries and transformations
• decouple vendor-specific interfaces
• improve resilience

Data flows to implement

ERP to MES

• production orders
• item master and BOMs
• routings/recipes
• inventory availability
• due dates and priorities
• engineering change notifications

MES to ERP

• order completion
• production quantities
• scrap and rework
• material consumption
• labor reporting
• inventory movements
• quality results
• genealogy/traceability summaries

Equipment to MES

• machine status
• cycle counts
• alarms and downtime reasons
• temperature/pressure/speed values
• batch/lot execution data
• process confirmations

MES to Equipment

• recipe or parameter download
• job selection
• start/stop commands
• setpoint updates
• operator prompts or acknowledgements

Implementation best practices

Standardize master data

Make sure ERP and MES agree on:

• item numbers
• operations/routings
• units of measure
• lot/serial rules
• equipment IDs
• reason codes

Bad master data is one of the biggest causes of failed integrations.

Use event-driven messaging where possible

Instead of batch polling, use events such as:

• order released
• operation started
• machine faulted
• quality hold triggered
• order completed

This improves responsiveness and reduces integration complexity.

Define system ownership clearly

A common rule set:

• ERP owns commercial and planning data
• MES owns execution, WIP, and traceability
• Equipment/PLC owns real-time control
• Historian owns time-series process data

This avoids duplicate or conflicting records.

Build for store-and-forward and retries

Plant networks are not always reliable. MES integrations should handle:

• temporary outages
• duplicate messages
• delayed acknowledgements
• partial transactions

Keep latency requirements realistic

Not all shop-floor data needs sub-second response. Separate:

• real-time control, which belongs in PLC/SCADA
• near-real-time execution, which fits MES
• transactional business updates, which fit ERP

Common use cases

• production order dispatching
• recipe management and electronic batch records
• OEE and downtime tracking
• material consumption and genealogy
• quality inspections and nonconformance handling
• labor tracking and operator certification
• traceability for regulated industries

Example scenario

1. ERP creates a work order for 10,000 units.
2. MES receives the order and breaks it into line-level jobs.
3. MES downloads the correct recipe to the machine.
4. Equipment reports status and counts back to MES.
5. MES tracks scrap, downtime, and quality checks.
6. MES posts completion and inventory consumption back to ERP.

If you are designing this from scratch

Start with:

1. a clear process map
2. a master data alignment exercise
3. interface specifications for ERP↔MES and MES↔equipment
4. a pilot line or single value stream
5. exception handling and reconciliation logic

If you want, I can also provide:

• a reference architecture diagram
• a sample ERP/MES/PLC data model
• interface message examples
• a step-by-step integration roadmap