MES Software: Vendors, Features & Costs Compared 2026
MES software compared: vendors, functions per VDI 5600, costs (cloud vs. on-premise) and implementation. Honest market overview 2026.
By Christian Fieg · Last updated: April 2026
What is rework management in manufacturing?
Rework management is the systematic process of recording, analysing, and reducing rework in production. Rework means any additional operation performed on a unit that did not pass inspection the first time but can be corrected and still shipped as a good part.
The formula is simple: Rework rate = (Reworked units ÷ Total units produced) × 100.
If a plant produces 10,000 units per day and 350 require rework, the rework rate is 3.5 %. That number looks small in a percentage. Translated into cost, it is not. Those 350 units consume repair labour, tie up a repair station, slow throughput, and carry a higher risk of field failure than first-pass-good units.
Most plants know their scrap rate. Few know their rework rate with the same precision, because rework is harder to track. A scrapped part disappears. A reworked part re-enters the production flow and looks like a good part from that point forward. Without an MES that logs the rework event, links it to the unit's serial number, and traces it back to the originating station, rework becomes invisible. It inflates labour costs, consumes capacity, and hides the true cost of quality problems.
What does rework actually cost?
Rework cost is almost always underestimated because the visible cost (repair labour) is only one component. The full cost includes several hidden layers:
| Cost component | What it includes | Typical magnitude | Visibility |
|---|---|---|---|
| Direct repair labour | Technician time at the repair station: disassemble, fix, reassemble, retest. | €5–€50 per event for components. €50–€500 for complex assemblies. | Visible. This is the number most plants track. |
| Repair material | Replacement components, solder, adhesives, cleaning agents consumed during repair. | €1–€20 per event. | Partially visible. Often booked to "consumables," not to the defect. |
| Lost throughput | If the repair station is inline, it creates a bottleneck. If it is offline, the unit still occupies WIP space and delays delivery. Either way, the plant produces fewer shippable units per hour. | Often the largest cost. A bottleneck machine running at €800/hour that loses 5 % capacity to rework loses €40/hour, or €640/day on two shifts. | Invisible without MES. The throughput loss is absorbed into "normal" production time. |
| Inspection and retest | Every reworked unit must be retested. Retest consumes inspection capacity, gauge time, and sometimes destructive test samples. | €2–€30 per event depending on test complexity. | Often invisible. Retest is performed by the same inspectors and counted as "inspection," not as rework cost. |
| Increased field failure risk | Reworked units have a statistically higher field failure rate than first-pass-good units. The repair may fix the symptom but not the underlying weakness. If the unit fails at the customer, the warranty cost is 10–100× the internal repair cost. | Hard to quantify per unit. But at population level, a plant with 5 % rework rate will have a measurably higher warranty RPT than a plant with 1 % rework rate. | Invisible until the warranty claims arrive, months later. |
| Administrative overhead | Documenting the defect, logging the repair, updating the traceability record, generating the rework report. In plants without MES, this is done on paper or not at all. | €2–€10 per event in manual documentation time. | Invisible. Absorbed into "paperwork." |
Rule of thumb: The true cost of rework is 3–5× the visible repair labour cost. A plant that tracks €50,000/month in repair labour is likely spending €150,000–€250,000/month on rework when all hidden costs are included.
When should you rework, and when should you scrap?
Not every defective unit should be reworked. The decision depends on economics, quality risk, and customer requirements:
| Factor | Rework | Scrap |
|---|---|---|
| Cost | Rework cost (labour + material + retest) is significantly less than the unit's material value. | Rework cost approaches or exceeds the unit's material value. Scrap and make a new one. |
| Quality risk | The repair fully restores the unit to specification. No residual quality risk. | The repair cannot guarantee specification compliance, or the reworked unit has a higher field failure probability. |
| Customer requirements | The customer accepts reworked units (most do, if documented and retested). | The customer explicitly forbids rework (some automotive OEMs and aerospace customers do for safety-critical parts). |
| Traceability | The rework can be fully documented and traced (which unit, which defect, which repair, which retest result). The MES enables this. | The rework cannot be documented because there is no traceability system. A reworked unit without documentation is a liability. |
| Capacity | The repair station has available capacity. Rework does not create a bottleneck. | The repair station is at capacity. Adding more rework slows the entire line. Scrap is faster. |
The MES makes this decision data-driven rather than gut-based. It shows the rework cost per defect type, the repair station utilisation, and the first-time-fix rate. If a specific defect type has a 60 % first-time-fix rate (meaning 40 % of repaired units fail retest and need a second repair), the economics shift towards scrap. The MES provides the data to see this.
How does rework affect First Pass Yield and OEE?
Rework sits in the gap between two metrics that most plants track separately but rarely connect:
| Metric | How rework affects it | The problem this creates |
|---|---|---|
| OEE Quality rate | OEE Quality = Good parts ÷ Total parts. A reworked unit that passes retest counts as a good part. So OEE Quality does not capture rework. A plant can have 99.5 % OEE Quality and still rework 5 % of its output. | OEE looks good. Management is satisfied. But the plant is spending €150,000/month on hidden rework cost that OEE does not show. |
| First Pass Yield (FPY) | FPY = Units good on first attempt ÷ Total units. FPY captures rework: a reworked unit failed the first attempt. FPY = 100 % minus (scrap rate + rework rate). The same plant with 99.5 % OEE Quality might have 94.5 % FPY. | FPY reveals the true quality picture. But most plants do not track FPY because it requires knowing which units were reworked, not just how many were shipped. |
This is why rework management matters. OEE hides rework. FPY reveals it. The MES tracks both: OEE from cycle counts (the production metrics module), and FPY from rework station logs that link each repair event to the original unit and the originating station.
How does an MES enable rework management?
| MES capability | How it supports rework management | Without MES |
|---|---|---|
| Rework event logging | The operator scans the unit at the repair station. The MES records: unit ID, defect type, originating station, repair action performed, technician, timestamp, repair duration, retest result (pass/fail). | Rework is logged on paper tags (if at all). The data is never aggregated. Nobody knows how many units were reworked last month, let alone which station caused the defects. |
| Origin tracking | The MES links each rework event to the station where the defect was created, not where it was detected. This is the critical distinction. A solder defect detected at end-of-line inspection originated at the wave solder station. The MES traces it back. | The repair technician fixes the defect but nobody investigates where it came from. The same defect repeats tomorrow from the same station. |
| Rework Pareto | The MES generates a Pareto automatically: defect type × frequency × originating station × cost. The top 3 defect types typically account for 60–80 % of all rework. At Neoperl, this type of quality data analysis contributed to 15 % scrap reduction. | The quality team knows "there is a lot of rework" but cannot say which defect to attack first. Improvement efforts are scattered across many small issues instead of focused on the vital few. |
| Process parameter correlation | The MES process data module links rework events to the process parameters at the originating station. "Units reworked for cold solder were produced when wave solder temperature dropped below 248 °C." Root cause identified. Corrective action: tighten temperature control. | Process parameters and rework data live in separate systems. The correlation exists in the physics, but nobody can see it because the data is not connected. |
| Routing enforcement | The MES prevents reworked units from skipping retest. A digital Poka Yoke: the unit cannot proceed to packing until the retest station confirms a pass. No exceptions, no shortcuts. | Reworked units occasionally bypass retest when the line is under pressure ("it looks fine, just ship it"). Those are the units that generate warranty claims. |
FAQ
What is the difference between rework and repair?
In most manufacturing contexts the terms are used interchangeably. Strictly, ISO 9000:2015 distinguishes them: rework makes a nonconforming product conform to requirements (the output meets spec). Repair makes a nonconforming product acceptable for intended use but does not fully meet original requirements (the output is "use as is" with a concession). The distinction matters for quality documentation: a reworked unit is conforming, a repaired unit is non-conforming-but-accepted. The MES should track the disposition (rework vs. repair vs. scrap) separately because the implications for traceability and customer reporting are different.
Can rework be eliminated entirely?
In theory, yes. In practice, the question is economics. Eliminating the last 0.5 % of rework may require capital investment (new tooling, automation, vision inspection) that costs more than the rework it prevents. The rational approach: use the rework Pareto to attack the top defect types first. Each PDCA cycle eliminates the largest remaining source of rework. At some point the remaining rework is so low-volume and varied that further reduction is not cost-effective.
How does rework management relate to RPT?
Internal RPT (Repairs Per Thousand) is the direct metric of rework volume: RPT = (Reworked units ÷ Total units) × 1,000. Rework management is the process that reduces RPT. The MES provides the RPT number and the drill-down data (which defects, which stations, which shifts) that the rework management process uses to identify and eliminate root causes.
What is a "rework loop" and why is it dangerous?
A rework loop occurs when a repaired unit fails retest and returns to the repair station for a second (or third) attempt. Each loop iteration consumes repair labour, occupies the repair station, and degrades the unit further. The MES tracks the loop count per unit. If a unit fails retest twice, the disposition should change from "rework" to "scrap." Plants without MES tracking often let units loop 3, 4, 5 times, burning repair capacity on units that will never pass. Monitoring the first-time-fix rate at the repair station stops this waste.
Related: RPT · PPM · Root Cause Analysis · Poka Yoke · Quality Control · Six Sigma · OEE Explained · SYMESTIC Production Metrics · SYMESTIC Process Data · MES: Definition & Functions
About the author
Christian Fieg
Head of Sales at SYMESTIC. Six Sigma Black Belt. At Johnson Controls, managed the MES and traceability systems that tracked every rework event across 30+ manufacturing processes and 900+ machines. Learned that the plants with the lowest rework cost were never the ones with the busiest repair stations, but the ones that made the repair station unnecessary. · LinkedIn
Start working with SYMESTIC today to boost your productivity, efficiency, and quality!
