What is a manufacturing execution system (MES)? The complete guide

Somewhere on your shop floor right now there is a job nobody can fully account for. The material was issued against it — probably — but the store keeper wrote the quantity on a slip that is now under a coffee cup. Two operations are done, one is half-done, and the reject count lives in the line supervisor's head. When planning asks how many good pieces you actually have, the honest answer is "let me go and count." That gap — between what the plan assumes and what the floor actually did — is the problem a manufacturing execution system exists to close.

This guide is written for the person evaluating the software category — a plant head, production manager, or business owner deciding what to buy and what to ask vendors. It covers what the software actually does step by step, the capabilities that separate a real production system from a spreadsheet with a login screen, and how Fast Production Software implements each one. If you want to learn the underlying discipline first — the practice of production management itself, independent of any tool — start with our pillar guide, What is production management software?, and come back here when you are ready to compare software.

1. What is a manufacturing execution system?

A manufacturing execution system — also sold as production management, production control or shop-floor software — gives a plant a single, structured way to release an engineering definition, open work orders, issue material, execute operations, capture good and reject output, transfer finished goods, and record rejection and rework. It sits between planning and inventory: planning tells it what and how much to make, inventory holds the material it consumes, and quality inspects what it produces.

The starting point is always the same three-part engineering definition: a Bill of Materials (what the product is made of), a Bill of Resources (the machines, labour and tools it takes to make it), and a process or route sheet (the ordered sequence of operations through each work center). Once that is released, the system can explode it into a material requisition and a route, and every downstream movement becomes a numbered, owned, dated document rather than a note on a job card.

The key word is executed. Plenty of tools can store a BOM. The difference a real production system makes is that nothing about the execution is optional or memory-dependent:

• Every job is a work order — an instruction to make a quantity of an item by a date, born as a draft and released before the floor touches it, not a verbal "start on those brackets"
• Every issue is recorded — material leaves store stock against a specific work order, so consumption can be compared to the BOM instead of guessed at
• Every operation has output — good and reject quantity are booked at each step, so work-in-progress is a number, not an estimate
• Every finished piece is transferred — completed output lands back in inventory as stock at a defined transfer point, ready to dispatch
• Every reject gets a route — parts rejected on the line escalate into a controlled rework loop or are scrapped with a record, instead of piling up in a red bin nobody reconciles

For organisations working to a quality system such as ISO 9001 or IATF 16949, this is not just good housekeeping — it is the working machinery behind traceability: which material lot went into which work order, who ran which operation, what was rejected and what was reworked. Our material issue & WIP and FG, rejection & rework pages cover the traceability angle in detail.

2. Why spreadsheets and paper job cards fail

Most plants do not start with no system. They start with a planning spreadsheet, a printed job card per work order, and a store keeper's issue register. It feels adequate until volume, staff turnover or a customer audit exposes it. The failure modes are consistent enough to list:

• No single record. The work order lives on a printed card, the material issue lives in the store register, the reject count lives on the line, and finished-goods receipt lives in inventory's book. Reconstructing what one job actually consumed takes an afternoon.
• Stock that lies. A spreadsheet cannot tell the difference between material that is reserved for a job and material that has actually been issued. So on-hand looks fine until the second work order tries to draw the same steel that was quietly committed to the first.
• Invisible WIP. Between "issued to line" and "finished goods received", the quantity is nobody's number. Half-machined parts sit at a work center and no report can say how many good pieces are three operations deep.
• Reject that evaporates. Parts fail at an operation, go into a bin, and the loss never lands against the work order. Yield looks better on paper than in the stores, and the same defect returns next month because nobody counted it the first time.
• No pattern visibility. Ten rejections at the same work center look like ten separate bad days. Without capturing reject by operation and by defect, a systemic tooling problem hides inside individual anecdotes.
• Audit pain. When a customer auditor asks to trace a lot from finished goods back to the raw-material heat number, a stack of job cards is not a record. Assembling the chain retroactively is miserable — and visible.

None of this fails loudly. That is what makes it dangerous: a spreadsheet never sends you a report saying material is double-committed, WIP is ageing at the grinding section, and reject at one work center has quietly doubled.

3. What the software does — the production lifecycle

Whatever the vendor, discrete production follows broadly the same sequence. What the software does is make each step explicit, owned and recorded — so the process runs the same way whether the floor is running five work orders a week or five hundred.

The lifecycle is a loop, not a line — reject captured at each operation feeds the rework decision, and salvaged parts flow either to finished goods or back into the main work-order flow.

4. Core capabilities checklist

Feature lists blur together quickly. These are the eight capabilities that actually determine whether a product can run a shop floor — use them as your evaluation checklist.

Alongside these eight, check the commit discipline: does the system understand that stock only truly moves at issue and at FG transfer — not at reserve, not at WIP booking? A product that deducts stock at the wrong moment will double-count on-hand and misstate WIP, and no amount of dashboards will fix a ledger that commits at the wrong point.

5. MES vs ERP — and how it connects to planning, inventory and quality

Buyers often discover mid-evaluation that "production software" and "ERP" are answering two different questions. They are related but not the same, and the best arrangement runs both on one platform rather than bolting a shop-floor tool onto a separate accounting system.

The point is not that one replaces the other — it is that they must share the same data. A manufacturing execution system is a consumer of planning and inventory and a feeder of quality and finished-goods stock:

• Planning / MRP tells production what to make. Work orders are generated from the sales or component plan, and MRP reservation creates the earmarks production issues against. See Fast Planning / MRP.
• Inventory / WMS holds the raw material. Every material issue and every FG transfer posts to the same stock ledger the warehouse uses, so raw-material depletion and finished-goods receipt are immediately visible to inventory and dispatch. See Fast Inventory & Quality.
• Quality / QMS inspects what is produced. Operation status branches into in-process and finished-goods inspection, and line rejections and rework decisions flow into the quality module's non-conformance and rework handling.

In Fast Production these are not integrations in the brittle sense — they are the same platform and the same document engine, so a work order, a material issue, a WIP posting and an FG transfer all interoperate natively with warehouse stock and quality inspection. Nothing is re-keyed at a boundary.

6. Who needs a manufacturing execution system?

Not every workshop needs a dedicated system on day one. These are the situations where the spreadsheet-and-job-card approach reliably stops being enough:

The practical trigger is usually one of three events: a job that consumed far more material than the BOM said and nobody could explain where it went; a customer audit finding on traceability or on reject control; or a growth stage where the number of concurrent work orders outruns the memory of the one planner who used to hold it all together.

7. How to evaluate a production system

Most demos look good. The differences show up in the specifics, so evaluate against your own shop floor rather than the vendor's script.

8. How Fast Production Software implements each capability

Fast Production Software is the shop-floor manufacturing-execution product of the Fast Suite, built in Pune by Improsys under the Fast Technology brand. It runs each capability above with real, named screens — the same ones you would see in a demo:

9. Frequently asked questions