What Is MES? Manufacturing Execution Systems and ISA-95 for Automation Engineers

A manufacturing execution system (MES) is the plant software that sits between business planning (ERP) and the control layer (PLCs, SCADA and DCS). It turns orders into dispatched work, records what was actually made, from which materials and to what quality, and measures performance such as OEE. ISA-95 is the standard that describes where it fits.

Where MES sits: the ISA-95 levels

ISA-95, published internationally as IEC 62264, is the standard for integrating enterprise systems with control systems. Its functional hierarchy gives each layer a job and a time scale, and it is the vocabulary every MES project uses.

Level 3 is MES territory. ISA-95 calls it manufacturing operations management and divides it into four areas: production, maintenance, quality and inventory. Products sold as MES usually cover production and part of quality. Maintenance often lives in a separate CMMS, laboratory results in a LIMS, and process values in a historian. On a real site, "the MES" is often several of these working together.

The reason the layer exists is a gap. ERP knows the plant must make 5,000 units of a product by Friday. The PLC knows how to run a filler. Nothing in between knows which line will run it, which lot of raw material went in, which operator was on shift, or whether the in-process checks passed. MES fills that gap.

What a manufacturing execution system actually does

Most systems offer some combination of these functions:

• Dispatching. Work orders are scheduled to lines and shifts and released to operators in order.
• Data collection. Counts, machine states, downtime and reject reasons arrive from PLCs and SCADA, and operators add what the machines cannot know.
• Genealogy and traceability. Every lot consumed is linked to every lot produced, so a recall can be traced forwards and backwards in minutes rather than days.
• Quality. In-process checks, sample results, statistical process control and the ability to put material on hold.
• Performance. OEE, downtime Pareto charts and shift reports.
• Documents and recipes. The right work instruction and the right recipe version reach the right machine.
• Labour and resources. Who worked on what, and whether they were trained to do it.

OEE: where most MES projects start

In this lesson by CQE Academy, from our free Power BI OEE course, overall equipment effectiveness is explained and calculated. The worked example below uses the same three ratios with numbers from a packing line.

Overall equipment effectiveness multiplies three ratios: availability, performance and quality. A worked example with illustrative numbers:

• A shift has 420 minutes of planned production time. The line stops for 42 minutes, so it runs for 378. Availability is 378 / 420 = 90%.
• The ideal rate is 60 units a minute, so 378 minutes could make 22,680 units. The line made 20,412. Performance is 20,412 / 22,680 = 90%.
• Of those, 19,391 were good. Quality is 19,391 / 20,412 = 95%.
• OEE is 0.90 x 0.90 x 0.95, about 77%.

Every number in that calculation starts in a PLC. An MES can only be as accurate as the tags it reads: a machine state, a good-part counter, a reject counter and a stop reason code. Automation engineers who design those tags cleanly, with one state tag per machine and agreed reason codes, make or break MES projects. The Lean, OEE and TPM course linked below teaches the six big losses behind the three ratios.

How MES talks to the control layer

The connection runs in both directions, and a good design treats it as a handshake rather than a stream of values.

The common technical routes are:

• [OPC UA](https://opcfoundation.org/about/opc-technologies/opc-ua/) from PLCs or SCADA servers, for structured data with quality and timestamps.
• [MQTT](https://mqtt.org/), increasingly used to publish machine data to a broker that MES and other systems subscribe to.
• SQL databases, where SCADA logs production data that MES reads, or where MES writes orders that SCADA reads.
• ERP interfaces built on the ISA-95 object models. B2MML, an XML implementation of those models originally published by MESA International, is one common format. MESA closed on 30 June 2026 and its content and educational programmes passed to ISA.

The handshake matters. When MES downloads a recipe, the PLC should confirm it received valid parameters before it runs, and MES should refuse to start an order the machine has not acknowledged. Most production-data errors trace back to a missing acknowledgement or a counter that reset without anyone noticing. Getting those signals clean and time-aligned is the subject of our guide to data engineering for industrial AI.

Buy, build, or both

Plants take three broad routes. Some buy a commercial MES product and configure it. Some build MES-style functions on a SCADA platform with SQL and scripting, which is common with platforms such as Ignition; our SCADA training guide explains that supervision layer. Smaller sites often start with a database, a few operator screens and a dashboard, and grow from there. The right route depends on the number of lines, how regulated the product is, and who will maintain it. Regulated plants add validation and electronic batch records on top, as our guide to pharmaceutical automation explains.

The commercial side is moving: industrial software is consolidating around the big automation suppliers, as Schneider Electric's planned acquisition of PTC shows, so check a product's roadmap as well as its feature list.

Whichever route is chosen, the same ISA-95 thinking applies: define equipment, materials, personnel and the production schedule once, and let every system use those definitions.

Skills an automation engineer needs for MES work

• PLC data design: states, counters, reason codes and handshakes that are unambiguous.
• SQL: almost every MES stores its data in a relational database, and reports are queries.
• ISA-95 vocabulary: work orders, segments, material lots and equipment hierarchy.
• Connectivity: OPC UA and MQTT, and how a historian stores time-series data.
• Reporting: turning raw counts into OEE and downtime dashboards people use.
• Process understanding: lean basics, so the numbers lead to changes on the line.

If you are building a wider plan, the 12-week automation curriculum shows where these fit after PLC and SCADA basics.

Free courses that build MES skills

EDWartens does not have a course called MES, and we would rather say so than stretch a title. These four free courses teach the parts an MES engineer uses every day:

• Ignition SCADA: gateway, tags, device connections, SQL history, reporting and MQTT.
• SQL for PLC and SCADA Engineers: the queries behind every production report.
• Power BI for Manufacturing: OEE and Production Dashboards: availability, performance and quality in DAX, and a downtime Pareto.
• Lean Manufacturing, OEE and TPM: the losses behind the numbers.

Every course is free in full: video lessons by independent YouTube creators (credited on each course page), written notes, a practice task per module and one final assessment of 15 questions with a 60% pass mark. A free account saves your progress; no card is needed.

The only paid item is the optional EDWartens Certificate of Completion. These four courses are beginner level, so it is a small one-off fee, US$8.99. It carries a code anyone can check at edwartens.com/verification. It is not a vendor or ISA certification and is not accredited. Create a free account to start.

Take the free course

Free

SCADA · Beginner · Free

Ignition SCADA by Inductive Automation

Free

Data and analytics · Beginner · Free

SQL for PLC and SCADA Engineers

Free

Data and analytics · Beginner · Free

Power BI for Manufacturing: OEE and Production Dashboards

Free

Manufacturing excellence · Beginner · Free

Lean Manufacturing, OEE and TPM