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 cycle time?
Cycle time is the time a process needs to complete one unit, measured from the start of one piece to the start of the next. It is a machine- and process-level metric. On a press line, it is the seconds between two stamped parts. On a filling line, it is the time between two filled bottles. On an assembly cell, it is the interval between two finished units leaving the station.
The definition sounds simple, but three distinctions matter in practice. Cycle time is different from takt time, which is demand-driven. It is different from lead time, which is customer-facing. And "actual cycle time" is different from "ideal cycle time", which is the benchmark used in the OEE Performance factor. Mix them up, and your KPIs lie — consistently and in the wrong direction.
This page explains the three definitions, how cycle time feeds into OEE, how it is measured automatically with an MES, and why the "ideal cycle time" on the machine nameplate is almost never the one that should drive your OEE calculation.
Cycle time vs. takt time vs. lead time
The three terms are the most frequently confused trio in production KPIs. The differences are precise, and each metric answers a different question.
| Metric | Definition | Driven by | Answers |
|---|---|---|---|
| Cycle time | Time between two consecutive units produced | Process & equipment | "How fast can we run this process?" |
| Takt time | Available production time ÷ customer demand | Customer demand | "How fast must we run to meet demand?" |
| Lead time | Time from order entry to delivery | Full value stream | "How long does the customer wait?" |
The healthy relationship is: cycle time ≤ takt time ≤ lead time. If cycle time exceeds takt time, the process cannot physically meet demand, no matter how much overtime you schedule. If cycle time is far below takt time, the line is overcapacity and overtime becomes overstock.
How to calculate cycle time
Two formulas dominate, and both are correct depending on what you measure.
Direct measurement (single unit):
Cycle time = End time of unit n − End time of unit n−1
Aggregated measurement (production run):
Cycle time = Net production time ÷ Good units produced
The direct measurement gives you the distribution — minimum, median, variance, outliers. The aggregated formula gives you a single average and is what most ERP systems calculate. The difference matters: an average of 32 seconds can hide the fact that 20 % of cycles take 50+ seconds, which is where your real performance loss lives.
| Cycle time variant | What it includes | Use case |
|---|---|---|
| Machine cycle time | Pure equipment cycle, no human work | Automated lines, OEE Performance factor |
| Operator cycle time | Manual work time per unit | Line balancing, ergonomics, labor planning |
| Total cycle time | Machine + operator + handling | Station design, takt balancing |
| Ideal cycle time | Theoretical best with zero losses | OEE benchmark, nameplate capacity |
Cycle time in the OEE calculation
Cycle time enters OEE through the Performance factor. The standard formula is:
Performance = (Ideal Cycle Time × Total Count) ÷ Run Time
The dangerous word in that formula is "Ideal". Three sources compete, and choosing the wrong one makes every OEE number you ever report structurally misleading.
| Source of ideal cycle time | Pros | Cons |
|---|---|---|
| Machine nameplate | Official, contractually agreed | Often measured in perfect lab conditions, not real production |
| Best observed cycle (e.g. 95th percentile of a good run) | Reflects what the process actually achieves when healthy | Needs historical data and an MES to compute reliably |
| Customer-contracted cycle | Directly tied to delivery commitments | May not match physical capability — OEE stops being a machine KPI |
The most common trap: plants use the nameplate cycle time for OEE because "that's what the machine builder said". When nameplate is 30 s and real best is 34 s, Performance can never exceed 88 %, no matter how well operations runs. Worse, when someone later "tunes" the ideal cycle time down to 34 s to look better, OEE jumps overnight without a single real improvement. That is one of the patterns covered in "OEE: One Number, Many Lies".
What an MES actually measures
Cycle time is cheap to define and expensive to measure correctly. Without an MES, most plants rely on one of three unreliable sources: operator tally sheets, ERP backflush divided by produced quantity, or standard planning values from the work plan. All three hide micro-stops, all three smooth out variance, and all three make performance loss invisible.
A modern MES changes the data quality in a specific way. It captures every single cycle via the machine signal — PLC trigger, OPC UA subscription or digital I/O pulse — and timestamps it to the second. Over a shift, that produces the full cycle time distribution per article per machine, not an average.
| Question | Without MES | With SYMESTIC MES |
|---|---|---|
| Average cycle time today | Available next shift, rough | Live on the dashboard, per article |
| How many cycles exceeded target by > 20 %? | Usually unknown | Counted automatically, trendable |
| Micro-stops vs. slow-running | Indistinguishable from scrap rate alone | Separated: short stops logged, slow cycles flagged |
| Real ideal cycle time for OEE | Nameplate (lies) or gut feeling | 95th-percentile of historical good runs, defensible |
At Meleghy Automotive, this shift from ERP-backflush data to per-cycle capture across six plants is what actually unlocked the 7 % output improvement. The lines were not broken. The measurement was.
How to reduce cycle time, in order of return
Not every cycle-time problem deserves an investment project. The return profile is steep: the first two levers below cost almost nothing and deliver most of the gain, the last two require capital and should only follow once measurement is clean.
| Lever | What it attacks | Typical return |
|---|---|---|
| Eliminate micro-stops | Cycles that run longer than target but aren't counted as downtime | 3–10 % Performance uplift |
| Reduce cycle-time variance | Operator dependency, tool wear, parameter drift | 5–15 % Performance uplift |
| SMED on changeovers | Lost cycles during setup | 30–90 % changeover reduction |
| Process re-engineering | Physical cycle limits (heat, pressure, chemistry) | Variable, project-specific |
| Automation / capex | Replacement of the bottleneck operation | 10–40 % cycle reduction, months of payback |
FAQ
Is a shorter cycle time always better?
No. A cycle time shorter than takt time means overcapacity, which creates overproduction, one of the seven wastes in Lean. The right target is: cycle time slightly below takt time, stable variance, and zero quality loss at that speed. Pushing cycle time below the process-capable limit also triggers scrap and micro-stops that destroy what you gained.
How is cycle time different from throughput?
Cycle time is time per unit. Throughput is units per time. They are mathematical inverses: Throughput = 1 ÷ Cycle Time. On a line producing 2,400 pieces/hour, cycle time is 1.5 seconds. Throughput is the customer-facing number; cycle time is the engineering-facing number.
What is a good cycle time?
The question is wrong. "Good" is not absolute — it is relative to the process's physical capability and the customer's takt. A 45-second cycle can be excellent on one machine and terrible on another. The useful benchmarks are: cycle time stability (low variance), cycle time vs. takt time (capacity headroom) and cycle time vs. ideal cycle time (Performance factor in OEE).
Why does the ERP-calculated cycle time differ from the MES-measured one?
ERPs calculate cycle time as Run Time ÷ Good Count, using backflushed quantities and shift-level time buckets. That blends cycle time with micro-stops, slow cycles and reporting lag. An MES measures each cycle at the PLC signal, second by second. Typical gap in brownfield plants: 8–20 %, always with ERP showing a rosier number. When plants first connect to SYMESTIC, this delta is one of the three most common "wait, really?" moments in week one.
Do I need an MES to track cycle time?
For a single line on a pilot basis, a data logger and Excel get you started. For anything beyond that — multiple articles per line, multiple shifts, multiple plants, automatic integration with OEE, order context from ERP — manual tracking collapses within weeks. The real decision is not "track cycle time yes/no" but "use an MES or keep pretending the ERP's cycle time is accurate". For most mid-sized manufacturers the answer settles itself within one audit cycle.
Related: OEE · Takt Time · SMED · Lean Production · Just-in-Time · Kaizen · Production KPIs
About the author
Christian Fieg
Head of Sales at SYMESTIC. Six Sigma Black Belt with 25+ years in automotive manufacturing, including global responsibility for MES and traceability across 900+ machines at Johnson Controls (China, Mexico, US, Europe). Author of "OEE: One Number, Many Lies". · LinkedIn
Start working with SYMESTIC today to boost your productivity, efficiency, and quality!
