What is MES?
MES is the system that executes and tracks on the shop floor the work order that ERP planned. It records which order was produced on which machine, with which operation, when and with what result.
In the ISA-95 model (internationally IEC 62264), MES sits in the manufacturing operations layer, between business planning (ERP) and the machines and control systems. Its typical functions are:
- dispatching work orders to the floor and tracking operations,
- recording downtime, production time and quantities,
- quality checks and scrap recording,
- material and batch traceability,
- performance indicators such as OEE.
trex MES: executing the work order on the shop floor
MES, ERP, APS, SCADA and PLC: which system does what in the factory?
In a factory, several production-related systems work in layers. The ISA-95 model separates these layers from business planning down to machine control; MES sits in the manufacturing operations layer in the middle. The diagram and table below show the task of each system and its relationship with MES; the position of APS is MindDX’s explanation.
| System | Main task | Relationship with MES |
|---|---|---|
| ERP | Plans what is to be produced, the material and the cost | Gives MES the work order and takes back the actuals |
| APS | Sequences jobs according to capacity | Gives MES the sequenced plan and updates it with the actuals from the floor |
| MES | Manages and records how work runs on the shop floor | Is the bridge between the plan and the shop floor |
| IIoT | Collects machine and sensor data | Supplies MES with data automatically |
| SCADA | Monitors and controls the process in real time | Can share process data with MES |
| PLC | Controls the machine directly | Is the source of status and counter signals |
- ERP · business planningOrders, bill of materials, routing, stock and cost.
- APS · schedulingWhich job, on which resource and when, based on finite capacity.
- MES · manufacturing executionExecuting the work order on the shop floor; recording quantities, times, downtime and quality.
- IIoT and SCADA · connectivity and monitoringCollecting machine and sensor data and monitoring the process.
- PLC and machines · controlThe machine itself and the controller that runs it.
The scheduling layer: the APS guide
How is real-time production data collected?
The source of the data depends on the age and type of machine. On machines with a controller, data is read directly; standard protocols such as OPC UA let machines from different makers speak the same language. On older machines without a controller, an added sensor or an IIoT module picks up the status signal. At manual workstations, the data comes from the operator’s entry at a terminal.
The data collected is usually machine status (running, stopped, faulted), good and rejected quantities, cycle time and process parameters. What makes it valuable is that it comes from the machine, at once, not by hand: you see what is happening now instead of a report written at the end of the shift.
Advantages and limits of real-time data collection
Data that arrives at once from machines and workstations lets the shop floor be run from the current record rather than from the end-of-shift report. Every advantage comes with a condition, though; most of what counts as a disadvantage stems not from the technology but from preparation. The table below sets both side by side for each topic.
| Topic | Advantage | Limit / watch out |
|---|---|---|
| Downtime | Stops are seen at once, together with their cause; the response does not wait for the end of the shift. | Picking up the machine signal requires investment in sensors, connection modules and the shop-floor network. |
| Cycle time | The actual cycle time is measured; planning and costing rest on records instead of assumptions. | On older machines, access to the controller may be limited; the data is taken through an added module or an indirect signal. |
| Data entry | Quantities and times written by hand decrease, and so do errors made when copying from paper to a spreadsheet. | Operator acceptance is needed: the terminal screen and the downtime reasons should be designed together with the shop floor. |
| Traceability | Batch, machine, operator and material come together in one record; in a quality problem the trail can be followed backwards. | Data quality: if signals are not matched correctly to the work order, wrong records multiply just as fast; validation rules are needed. |
| OEE and indicators | OEE and loss analyses rest on evidence rather than estimates. | Without a process owner who interprets the data and ties it to decisions, the screens are watched but nothing changes. |
Common challenges in real-time data collection
Most challenges in data-collection projects are of a known kind and can be planned for in advance. Below are the most common ones and how each is usually handled.
| Challenge | How is it handled? |
|---|---|
| Older machines and different protocols | A machine inventory is taken; machines with a controller are connected through a standard protocol such as OPC UA, those without one through an added sensor or an IIoT module. |
| Shop-floor network and dropped connections | The shop-floor network is built for the production environment; when the connection drops, data is held on the device or in an intermediate layer and transferred once the connection returns. |
| Consistency of operator entries | Downtime reasons are defined briefly and in the language of the shop floor; an entry takes a few taps and is compared with the automatic signal. |
| Data quality, matching and master data | Machine, work-centre and item codes are matched in a single master-data list; validation rules are defined for outliers. |
| Work-order and item-code matching in ERP integration | The work-order number and the item code travel with the same key in both systems; unmatched records show up in a queue instead of disappearing silently. |
| Managing changes | A written process sets who updates what for a new machine, a new product or a new downtime reason; teams are trained on the experience of the first line. |
Machine connectivity and OPC UA: the data-collection section
MES and OEE: how is efficiency measured?
OEE (Overall Equipment Effectiveness) measures how efficiently a machine runs in its planned time through the components of availability, performance and quality. The calculation is simple; the hard part is collecting downtime, quantity and scrap data correctly and completely.
In records kept by hand, short stops and slow cycles often stay invisible, so OEE comes out better than it is. When the data comes from the machine through MES and IIoT, the loss becomes visible together with its cause, and improvement is aimed at the right place.
What is OEE and how is it calculated? Glossary entry
Product traceability and analysis of the collected data
Traceability is the ability to follow the history of a product or a batch backwards and forwards. At every operation, MES records which batch was processed on which machine, by which operator and with which material batch. In a customer complaint or a quality problem, the affected batches are then found from the record rather than by guesswork, and a recall can, if needed, be limited to those batches.
These records and analyses can run in an MES in the cloud or in one on the company’s own servers (on-premise); the choice depends on how reliable the connection between the shop floor and the data centre is, how much latency processes close to the machine can tolerate, and in which country and on whose infrastructure the data has to reside.
Typical analyses of the collected data are:
- a Pareto analysis of downtime reasons: which few causes make up most of the lost time,
- a comparison of shifts, lines and products: how the same work runs under different conditions,
- the relationship between quality and process parameters: how values such as temperature, pressure or speed relate to rejects,
- how cycle times change over time: updating the standard times used in planning and costing,
- artificial intelligence and machine learning, for example early signs of a fault or a quality forecast: these become meaningful only once the data is reliable, correctly matched and collected over a sufficient period; the basic analyses come first.
Lot and serial tracking: glossary entry
The difference between IIoT and MES
IIoT is the connectivity and data-collection layer: it connects machines and sensors to the network and gathers their signals. MES is the execution layer: it relates those signals to the work order, the operation, the product and the quality record.
Each is incomplete without the other. IIoT alone says «the machine stopped»; MES says which order stopped, for how long and why. That is why the two are often named as one category.
Integration with ERP
MES and ERP talk in both directions. Work orders, bills of materials and routings flow from ERP to the floor. Actual quantities, times, scrap and downtime flow back from the floor to ERP. Cost and delivery dates are then calculated from the shop-floor record, not from an estimate.
The precondition for integration is master data. If item codes, work centres and routings differ between the two sides, even the best connection quickly carries the wrong data. Consistent master data is therefore the first task of the project.
Prodaso: data exchange between the shop floor and ERP
Data security in MES
Because MES is connected to machines, its security is not only an IT matter: a weakness can affect production itself as much as the data.
The international family of standards for the cybersecurity of industrial automation and control systems is the ISA/IEC 62443 series. It defines requirements and processes for implementing and maintaining these systems securely, and treats security as a responsibility shared between the plant operator, product suppliers and integrators.
In practice, the basic measures are:
- separating the shop-floor network (OT) from the office network (IT) and limiting the traffic between them to the connections that are needed,
- role-based authorisation: operators, shift supervisors, quality staff and managers see only the screens and actions their work requires,
- update and patch management for servers, terminals and connection devices, within a maintenance window that does not stop production,
- remote-access control: supplier and support connections are opened on record, for a limited time and with approval,
- backups: configuration, master data and production records are backed up regularly, and restores are tested.
When is a production tracking system needed?
If several of the following signs are present, the reality of the shop floor reaches decisions late and incomplete:
- production quantities and downtime are written on paper or in a spreadsheet at the end of the shift,
- finding out what stage an order has reached on the floor means walking to the line,
- downtime reasons are unknown or entered by guesswork,
- the actual times and scrap in ERP do not match the shop floor,
- in a quality problem, which batch, which machine and which operator were involved cannot be traced.
What to look for when choosing MES software
Ask for a trial with your own machines and one of your own work orders, and ask these questions:
- Machine connectivity: how are machines of different makes and ages connected, and are standard protocols supported?
- Operator use: can the terminal screen be used quickly on the shop floor, and in Turkish?
- ERP integration: do work orders, routings and actuals flow through a documented interface?
- Coverage: are functions such as quality, traceability, maintenance and energy on the same record?
- Indicators: can OEE and downtime analyses be built on your own downtime reasons?
- Deployment and data: cloud or your own server, and in which region is the data held?
- Growth: is it possible to start with one line and extend to the whole plant?
What does the cost of an MES depend on?
There is no single price; the structure of the plant and the scope of the project determine the cost. When comparing quotations, ask about these items separately:
- the number of machines and workstations to be connected,
- the hardware and sensors needed to connect the machines,
- the functions to be covered: execution, quality, traceability, maintenance,
- the route and scope of ERP integration,
- operator terminals and the shop-floor network,
- training, go-live and subsequent support.
In which industries is MES used?
MES is not tied to a particular industry; it can be used wherever production runs on work orders. What changes is what is tracked most closely on the shop floor. The table below gives general examples; each company sets its priorities by its own product and its customers’ requirements.
| Industry | Typically tracked |
|---|---|
| Metalworking and machining | Machine-tool stops, cycle time, tool changes and set-up time, measurement results and scrap. |
| Plastic injection moulding | Cycle time per mould, mould changes, process parameters (temperature, pressure), rejects and scrap. |
| Food and beverage | Batch and lot tracking, best-before information, recipe compliance, quality-control records and line stops. |
| Automotive suppliers | Part and serial traceability, quality-control steps, quantities produced against customer call-offs, and downtime. |
| Electronics assembly | Assembly and test results by serial number, material batches, defect records and line balance. |
| Textiles | Machine stops, batch and colour tracking, quality-control results, scrap and output per shift. |
Where does a small business start with MES?
MES is not only for large plants. For a small or mid-sized business (SME), too, the right path starts not with a large project but at the point that causes the most trouble. A general roadmap:
- start with one line or with the bottleneck machine, and expand once results are visible there,
- collect downtime and produced quantities first; this can start with operator entries on a terminal or tablet, with automatic machine signals and sensors as a later step,
- pick a few simple indicators, for example downtime duration and reason, and use them in a daily meeting,
- grow step by step: add lines, quality records, traceability and maintenance as the need arises,
- keep the master record of orders and work orders in ERP; MES records what actually happens on the floor and reports it back to ERP.
Where to start
An MES project starts not with choosing software but with measuring the current state: which machines there are, where today’s data comes from, and which indicators will really be used in decisions.
MindDX answers these questions in Assess. Processes, the machine inventory and the relationship with ERP are reviewed. Which scope to start with, on which line and with which product, is written into the Validated Scope document. Often the recommended first step is to start with one line rather than the whole plant.