PRODUCTIVITY · CHRONIC LOSSES · OEE
Microstops: seconds that can add up to weeks of lost capacity.
What they are, why they often become accepted, how to calculate their cumulative impact, where they belong in OEE and which data a plant needs to investigate them as production losses.
KEY POINT
Annual impact = frequency × duration × machines × shifts × days
An individual event may be brief. The accumulated loss can still be substantial.
OPERATIONAL DEFINITION
The duration threshold for a microstop must be defined for the application.
In this guide, we use microstop as a practical term for a brief, recurring production interruption that is usually resolved quickly and, because of its short duration, may neither generate a maintenance work order nor receive a specific recorded cause. ISO 22400-2 sets no one-, five- or ten-minute threshold for defining a minor stoppage. The standard defines ADET (Actual Unit Delay Time; actual delay time of the unit) by the type of situation: interruptions caused by malfunction, minor stoppages and other unplanned intervals that extend order processing.
Use thresholds for escalation while retaining the measured loss.
For example, a plant may require a maintenance call or an operator-selected cause only for stoppages longer than a specified duration. That is an escalation rule. Excluding every event below that threshold from measurement is a separate decision. To understand lost capacity, retain at least the start time, duration, machine and context of each detectable interruption, even when events are later aggregated for reporting.
WHAT THE EVIDENCE SHOWS
Chronic losses are easily accepted and difficult to record.
Bengtsson, Alm and Tjulin distinguish sporadic losses—large, visible events that readily trigger escalation—from chronic losses: frequent, small interruptions often resolved by the operator. Chronic losses are particularly difficult to capture when OEE relies on manual or semi-automatic recording.
In the example, breakdowns totalled 11.25 eight-hour shifts, while microstops totalled 17.5 shifts. The authors constructed this theoretical example to illustrate accumulation; it is not a universal benchmark.
The same paper includes an industrial case at a discrete manufacturing plant in the automotive sector. Based on expert estimates, microstops accounted for approximately 319.4 hours and almost 40 lost shifts per year; more than 200 hours were unaccounted for in the OEE system. This result illustrates how materially the actual events can differ from system records; its value lies in that gap rather than extrapolation to every factory.
Surveys of 80 companies in 2001 and 71 in 2014 found persistent differences between the events managers considered disturbances and those actually recorded. The lack of a shared terminology makes systematic improvement harder.
A non-intrusive IIoT system was validated in CNC machining, mould machining and a beverage production line. In the latter, frequent microstops were resolved quickly and rarely appeared in formal records; automatic detection made previously unrecorded events visible.
ACCUMULATION CALCULATOR
How much time is lost when each stoppage lasts just a few seconds?
Enter your plant data. The tool calculates cumulative losses; recoverable hours and achievable savings require separate assessment. The actual potential depends on root-cause analysis, a baseline and subsequent verification.
Effect of reducing frequency
CLASSIFICATION WITHIN OEE
ISO 22400 and classical TPM can assign the same microstop to different components.
In the published ISO 22400 formula, APT (Actual Production Time; actual production time) includes only the time during which the unit is producing and performing value-adding functions. If a microstop occurs within PBT (Planned Busy Time; planned busy time) and no production takes place during that interval, APT decreases and the loss appears under Availability. ISO defines ADET specifically for malfunction-related interruptions, minor stoppages and other unplanned intervals.
In Nakajima’s classical TPM approach, idling and minor stoppages are generally grouped with speed losses and reflected in Performance. Schiraldi and Varisco show that comparing ISO 22400 with the historical TPM formulation requires explicit assumptions and machine-state definitions.
A = 450 / 480 = 93.75%
E = 100 %
Q = 100 %
A = 100 %
Performance = 450 / 480 = 93.75%
Q = 100 %
This example demonstrates an aggregation property: the final OEE can be identical even when the component assigned the loss changes. In an actual plant, quantities, states and other losses interact. The original event must therefore be retained independently of the reporting convention.
Practical recommendation
Detect the event, retain its duration and cause, then apply the company’s documented OEE rule. Use the microstop threshold for the agreed operational purpose while keeping all detectable events. This allows the reporting convention to change without losing the machine’s actual history.
DETECTION AND DIAGNOSIS
Duration quantifies the loss. The cause identifies what needs attention.
| Cause category | Examples | Useful data | Potentially responsible team |
|---|---|---|---|
| Material supply | Empty hopper, bridging, interrupted dosing, missing component. | Level, feeder signals, recipe, batch, timestamp. | Production, logistics, materials, maintenance. |
| Sensors and automation | False trigger, contamination, misalignment, alarm reset. | PLC alarms, digital input, reset count, environmental conditions. | Automation, maintenance, production. |
| Mould / tooling | Stuck part, irregular ejection, slide issue, mould sensor. | Cycle, cavity, ejection signals, pressure/position, product. | Tooling, process engineering, maintenance. |
| Peripherals | Robot, dryer, temperature control unit, conveyor, dosing unit. | Peripheral state, alarms, handshake, temperature or flow. | Maintenance, automation, process engineering. |
| Human interaction | Repositioning a part, clearing an obstruction, acknowledging an alarm, waiting for relief. | Machine event + contextual operator entry. | Production, ergonomics, training, engineering. |
| Line flow | Downstream blockage, upstream starvation, full buffer. | Synchronised states from several machines or cells. | Production, internal logistics, line balancing. |
| Quality / process | Inspection, quick adjustment, parameter correction, removal of a suspect part. | Process variables, inspection, scales, vision system, recorded cause. | Quality, process engineering, production. |
The table supports investigation, with several possible causes for the same symptom. A microstop recorded only by duration remains a time loss; adding context can turn it into a testable hypothesis.
WHAT THE SYSTEM SHOULD RECORD
Improving a chronic loss requires an event history with sufficient detail.
Capture start and end automatically where technically feasible. A 20-second stoppage should be detectable even when nobody has time to type an entry.
Retain actual duration, using seconds rather than rounded minutes when data resolution permits.
Keep the technical state and OEE treatment separate from the cause code.
Machine, shift, product, mould, material, cavity, peripherals and relevant variables.
Pareto analysis by accumulated time and frequency. A 15-second event repeated 200 times may deserve more attention than a single 20-minute stoppage.
After intervening, compare frequency, accumulated duration and context using equivalent rules.
Automation should reduce administrative work.
Automatic event detection can reduce the need for operators to record every microstop manually. Where the available signals cannot establish the cause, the interface can request a brief classification when production resumes or group pending events for validation. Automatic time capture and human input on causes serve different purposes.
APPLICATION IN KERN IoP
Traceability makes a sequence of small interruptions a manageable production issue.
KERN IoP connects the machine’s physical event to its production context. Depending on the project, SMARTPRO can acquire cycle signals, states, alarms, sensors and personnel entries; MONITOR retains histories, classifies losses and provides access to frequency, duration and causes by machine, shift, product or period.
start, end, duration
automatic or validated
machine, mould, product
frequency + time
before / after
Some microstops are inherent to the process; others may require changes that are not economically justified. The aim is to measure first, identify recurring events, quantify their accumulation, investigate the dominant causes and then verify how much of the loss was reduced.
FREQUENTLY ASKED QUESTIONS
Microstops and production losses.
How long must a stoppage last to count as a microstop?
ISO 22400 sets no universal numerical threshold. A plant may use an operational threshold for escalation or cause recording. Document that threshold and retain the event in the measurement record.
Where are microstops accounted for in OEE?
In the published ISO 22400 model, if no value-adding production occurs during an interruption within PBT, APT decreases and the loss affects Availability. In classical TPM, minor stoppages are generally grouped under Performance. State the convention and apply it consistently.
Why do events lasting only seconds matter?
Annual impact depends on frequency and the number of machines, shifts and operating days. Bengtsson et al. present both a theoretical example and an industrial case where the annual sum of chronic losses was much larger than any individual event suggested.
Should the operator record every microstop manually?
Where a signal is available, automatic start and end detection is preferable. Operators can add the cause when it cannot be established from the machine signals, keeping the administrative burden proportionate.
Should the longest or the most frequent stoppages take priority?
Both views are needed. A Pareto chart of accumulated time identifies lost capacity; a frequency Pareto reveals instability and chronic events. A very frequent cause may justify intervention even when each event is brief.
How is the cost of a microstop calculated?
First calculate the accumulated time. Financial valuation can use attributable hourly costs or potentially unproduced units multiplied by a validated contribution margin. Check for possible double counting before combining cost models.
References
International Organization for Standardization. (2014). ISO 22400-2:2014. Automation systems and integration — Key performance indicators (KPIs) for manufacturing operations management — Part 2: Definitions and descriptions. ISO.
International Organization for Standardization. (2018). ISO/TR 22400-10:2018. Operational sequence description of data acquisition. ISO.
Bengtsson, M., Alm, P., & Tjulin, B. (2022). Visualizing the Effects of Chronic Versus Sporadic Losses in Manufacturing Industries – A Case Study. Advances in Transdisciplinary Engineering, 21, 3–14. https://doi.org/10.3233/ATDE220121
Bokrantz, J., Skoogh, A., Ylipää, T., & Stahre, J. (2016). Handling of production disturbances in the manufacturing industry. Journal of Manufacturing Technology Management, 27(8), 1054–1075. https://doi.org/10.1108/JMTM-02-2016-0023
Schiraldi, M. M., & Varisco, M. (2020). Overall Equipment Effectiveness: consistency of ISO standard with literature. Computers & Industrial Engineering, 145, 106518. https://doi.org/10.1016/j.cie.2020.106518
Hung, Y.-H., Li, L. Y. O., & Cheng, T. C. E. (2022). Uncovering hidden capacity in overall equipment effectiveness management. International Journal of Production Economics, 248, 108494. https://doi.org/10.1016/j.ijpe.2022.108494
Peças, P., de Oliveira, L. C., Febra, R., Duarte, S., Cruz-Machado, V., et al. (2026). Non-intrusive edge-based micro-stop detection using hybrid learning in manufacturing system. The International Journal of Advanced Manufacturing Technology, 145, 2309–2327. https://doi.org/10.1007/s00170-026-18438-2
TECHNICAL REVIEW
Dr.-Ing. Miguel Garzón · Technical Director, KERN IoP
Content reviewed by the technical team at KERN IoP, a division of PM Tec Engineering. Last reviewed: October 2026. Technical profile and publications →
Engineering
Resources
PM Tec Engineering, through its Industry 4.0 division, KERN IoP.
Cota, Cundinamarca · Colombia.
© 2026 PM Tec Servicios de Ingeniería SAS · Website updated in 2026.
You must be logged in to post a comment.