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 the PDCA cycle?
The PDCA cycle (Plan-Do-Check-Act) is a four-phase method for structured, iterative improvement. It originates from the work of Walter A. Shewhart in the 1930s and was popularised globally by W. Edwards Deming during his lectures in Japan in the 1950s. The concept is simple: define what you want to improve (Plan), test the change on a small scale (Do), measure whether the change worked (Check), and either standardise the improvement or adjust and try again (Act). Then repeat. The cycle never ends — that is the point. PDCA is the engine behind Kaizen (continuous improvement), it is the structural principle of ISO 9001 quality management systems, and it is the daily working method in any serious Lean or Operational Excellence programme. In manufacturing, the quality of the PDCA cycle depends entirely on the quality of the data feeding it. An MES provides the real-time production data — OEE, downtime reasons, cycle times, scrap rates, process parameters — that transforms PDCA from an opinion-based discussion into an evidence-based method.
What are the 4 phases of the PDCA cycle?
| Phase | Core question | What happens in this phase | MES data input | Common failure mode |
|---|---|---|---|---|
| Plan | What is the problem, what is the root cause, and what change will we test? | 1. Define the problem with data (not opinions). 2. Analyse the root cause (Ishikawa, 5 Why, Pareto). 3. Set a measurable target. 4. Design the countermeasure. 5. Define how you will measure success. | MES downtime Pareto, OEE trend, scrap rate per product/machine, alarm frequency analysis, cycle time distribution. At Neoperl, SPS alarm correlation identified the root causes that fed the Plan phase. | Skipping root cause analysis. Jumping from "OEE is low" directly to "buy a new machine" without understanding why OEE is low. |
| Do | Does the countermeasure work in practice? | Implement the countermeasure on a small scale — one machine, one shift, one product. Not a full rollout. This is a test, not a deployment. Document what was changed, when, and by whom. | MES captures the "during" data automatically: OEE, downtime, cycle times during the test period. No manual data collection required — the MES measures the same KPIs it always measures, now with the countermeasure active. | Implementing the change plant-wide before testing it. If the countermeasure fails, the entire plant is affected instead of one machine. |
| Check | Did the data confirm that the change achieved the target? | Compare before-data and after-data. Did the KPI improve? By how much? Is the improvement statistically significant or just random variation? Were there unintended side effects (e.g., availability improved but quality dropped)? | MES before/after comparison: OEE week-over-week, downtime minutes per shift, scrap rate per 1,000 parts. The SYMESTIC production metrics module provides these comparisons with timestamp precision — no manual spreadsheet work. | Declaring success after 2 days of data. Random variation looks like improvement over short periods. A credible Check phase needs at least 2–4 weeks of data, depending on the process. |
| Act | Should we standardise, adjust, or abandon the change? | Three possible outcomes: (a) the change worked → standardise it (update work instructions, train all shifts, roll out to other machines); (b) the change partially worked → adjust and run another Do-Check cycle; (c) the change did not work → abandon it, return to Plan with new hypothesis. | MES provides the ongoing monitoring after standardisation: if the KPI regresses weeks later, the MES detects it — the improvement was not truly standardised, and a new cycle is needed. | Skipping standardisation. The countermeasure works on day shift but is never documented or trained for night shift. The improvement disappears on the next shift change. |
Worked example: PDCA with MES data in manufacturing
A practical PDCA cycle on a press line, using real MES data at each phase:
| Phase | What the team does | MES data used |
|---|---|---|
| Plan | The MES downtime Pareto for press 5 shows that "material jam at infeed" accounts for 38 % of all downtime minutes in the last 4 weeks. The team performs a 5-Why analysis: material jam → strip misalignment → worn guide rail → no scheduled replacement interval for guide rail. Target: reduce "material jam at infeed" downtime by 50 % within 6 weeks. Countermeasure: replace guide rail every 4,000 operating hours (time-based preventive replacement). | MES downtime Pareto (reason code "material jam at infeed": 420 minutes / 4 weeks). MES operating hours counter for press 5: currently at 3,800 hours since last guide rail replacement. |
| Do | Guide rail replaced on press 5. Maintenance schedule updated: replace every 4,000 hours. The change is applied to press 5 only — not yet to the other 4 presses in the line. The team documents the change with a date stamp and photo. | MES continues recording downtime reasons and operating hours for press 5 automatically. No additional manual data collection is needed. |
| Check | After 4 weeks, the MES shows: "material jam at infeed" downtime on press 5 dropped from 420 minutes (prior 4 weeks) to 95 minutes (post-change 4 weeks). That is a 77 % reduction — exceeding the 50 % target. No negative side effects: scrap rate unchanged, cycle time unchanged. | MES downtime Pareto for press 5, filtered to the 4 weeks after the guide rail replacement. Before/after comparison generated directly in the SYMESTIC dashboard. |
| Act | The countermeasure worked. Standardise: (a) update the maintenance plan for all 5 presses — guide rail replacement every 4,000 operating hours; (b) train maintenance team on the new interval; (c) set MES operating-hour alert at 3,800 hours to trigger the replacement before the next jam occurs. The MES monitors whether "material jam at infeed" stays below 100 minutes per 4-week period going forward. | MES operating-hour counter per machine with threshold notification. Ongoing MES downtime monitoring: if the KPI regresses, the PDCA cycle restarts automatically. |
This is PDCA in its purest form: a specific problem, measured with data, solved with a targeted countermeasure, verified with data, standardised with data, and monitored with data. Without the MES, the same cycle would rely on shift supervisor estimates ("I think we had fewer jams this month"), which is unreliable and non-reproducible.
How does PDCA compare to DMAIC?
PDCA and DMAIC (Define-Measure-Analyse-Improve-Control) from Six Sigma are both structured improvement methods. They are not competitors — they operate at different levels of complexity:
| Dimension | PDCA | DMAIC |
|---|---|---|
| Origin | Shewhart (1930s), popularised by Deming (1950s) | Motorola (1986), popularised by GE under Jack Welch (1990s) |
| Typical cycle duration | Days to weeks. Fast iterations. | Weeks to months. Deep statistical analysis. |
| Statistical rigour | Low to moderate. Pareto, trend charts, before/after comparison. | High. Hypothesis testing, regression, design of experiments (DOE), process capability analysis (Cp/Cpk). |
| Who runs it | Shop floor teams, shift leaders, CI facilitators. Does not require specialised statistical training. | Green Belts, Black Belts with statistical training. Requires dedicated project resources. |
| Best suited for | Problems with an identifiable root cause that can be solved with a single countermeasure. "The guide rail wears out and causes jams." | Complex, multi-variable problems where the root cause is not obvious. "Scrap rate varies between 2 % and 8 % and we don't know why." |
| Relationship to MES | MES provides the Plan and Check data: downtime Pareto, OEE trend, before/after comparison. | MES provides the Measure and Analyse data: process parameter history, cycle-level data exports for statistical analysis, SPC charts. |
In practice, most manufacturing plants use PDCA for 80 % of their improvement activities (shopfloor Kaizen events, daily CI work) and DMAIC for the remaining 20 % (complex, cross-functional projects with dedicated Six Sigma resources). The MES feeds both: standard dashboards for PDCA, detailed data exports for DMAIC.
Why do most PDCA cycles fail in practice?
The PDCA concept is simple. The execution fails for predictable reasons — all of which are data problems:
| # | Failure mode | What happens | How MES data prevents it |
|---|---|---|---|
| 1 | Plan without data | The team selects a problem based on the loudest complaint, not the biggest loss. The root cause analysis is a 10-minute brainstorm instead of a data-driven investigation. The target is vague: "improve OEE." | MES downtime Pareto objectively ranks losses by impact. The biggest loss is a fact, not an opinion. At Meleghy, MES data drove a 10 % reduction in downtime — because the PDCA cycles addressed the actual top losses, not the perceived ones. |
| 2 | Do without containment | The countermeasure is rolled out to all machines simultaneously. If it does not work — or makes things worse — the entire plant is affected. | MES allows machine-level before/after comparison. Test on one machine, measure the result, then roll out. The MES dashboard shows the test machine versus the control machines in real time. |
| 3 | Check without measurement | "I think it's better" is not a Check phase. Without before/after data, the team cannot distinguish real improvement from random variation or confirmation bias. | MES provides timestamped before/after comparisons automatically. The Check phase becomes: "downtime for reason code X was 420 min in the 4 weeks before and 95 min in the 4 weeks after." That is a Check. |
| 4 | Act without standardisation | The improvement works — but only the day shift knows about it. Night shift reverts to the old method. Within 2 weeks, the KPI is back to the baseline. | MES ongoing monitoring: if the KPI regresses, the MES detects it in the next shift's data. The regression triggers a new investigation — not 3 months later when someone notices in the monthly report, but the next day. |
| 5 | No second cycle | The team completes one PDCA cycle, declares victory, and moves on. Six months later the problem returns — because the root cause was only partially addressed, or conditions changed. | The MES makes PDCA cyclical by design: the continuous monitoring of the same KPIs means that any regression or new loss pattern is visible immediately. The data itself triggers the next Plan phase. |
FAQ
What is the difference between PDCA and PDSA?
Deming himself preferred PDSA (Plan-Do-Study-Act) over PDCA. He argued that "Study" implies deeper analysis and understanding of the results, while "Check" suggests merely inspecting whether a target was met. In practice, PDCA and PDSA describe the same iterative process. ISO 9001:2015 uses the PDCA terminology explicitly in its clause structure (clauses 4–10 map to PDCA). Most manufacturing organisations use PDCA. The difference is semantic, not methodological — what matters is whether the Check/Study phase is done rigorously with data, regardless of which letter is used.
How does PDCA relate to Kaizen?
Kaizen is the philosophy — the belief that continuous small improvements, sustained over time, produce better results than occasional large projects. PDCA is the method — the structured way to execute each individual Kaizen improvement. Every Kaizen event, every A3 problem-solving sheet, every improvement suggestion from the shopfloor follows the PDCA logic: define the problem (Plan), test a solution (Do), verify the result (Check), standardise or adjust (Act). Without PDCA, Kaizen has no structure. Without Kaizen, PDCA has no organisational commitment. They are complementary, not interchangeable.
How does PDCA appear in ISO 9001?
ISO 9001:2015 is explicitly structured around the PDCA cycle. Clause 4 (Context) and Clause 5 (Leadership) set the frame. Clause 6 (Planning) = Plan. Clause 7 (Support) and Clause 8 (Operation) = Do. Clause 9 (Performance evaluation) = Check. Clause 10 (Improvement) = Act. This is not a loose analogy — the ISO standard's Annex states: "The PDCA cycle can be applied to all processes and to the quality management system as a whole." Any organisation with ISO 9001 certification is, by definition, committed to PDCA — even if the shopfloor team does not call it that.
How many PDCA cycles should a plant run simultaneously?
It depends on the organisation's capacity to execute, not on ambition. A common mistake is to start 15 PDCA cycles simultaneously, complete none of them, and declare that "PDCA doesn't work here." A practical guideline: 2–3 active PDCA cycles per production area at any time. Each with a clear owner, a defined target, and a timeline. The MES supports this by making the status visible: which KPIs are under active improvement, which are stable, which are regressing. At Neoperl, SYMESTIC was implemented as a KVP-Werkzeug (CI tool) — the MES data directly fed the prioritisation of which PDCA cycles to start next.
Related: Kaizen · Six Sigma · Lean Management · Operational Excellence · ISO 9001 · Shopfloor Management · OEE Explained · SYMESTIC Production Metrics · SYMESTIC Alarms Module · MES: Definition & Functions
About the author
Christian Fieg
Head of Sales at SYMESTIC. Six Sigma Black Belt. Ran PDCA and DMAIC cycles on 4 continents at Johnson Controls — where the cycles that worked always started with data and the cycles that failed always started with opinions. Author of OEE: Eine Zahl, viele Lügen. · LinkedIn
Start working with SYMESTIC today to boost your productivity, efficiency, and quality!
