Machine downtime is the first source of availability loss in industrial production. Yet on many sites stops are noted casually, in notebooks or Excel files, without real structure. Their qualification stays informal and their use stays partial. The most costly losses remain hidden in indicators. Capturing standardized downtime reasons at the source is one of the fastest levers to stabilize OEE and drive real actions.
Long stop, micro-stop, unqualified stop
A production stop is any interval where a machine or line is not in effective production when it should be. The concept covers very different realities: breakdowns, material waits, changeovers, cleaning operations, micro-stops, degraded modes. Two broad families stand out. Long stops (above a few minutes) are visible and usually trigger an immediate maintenance response. Micro-stops (a few seconds to a few minutes) stay invisible individually but cumulate quickly. They often represent the largest cumulative OEE loss of the week. A third category weighs heavily: unqualified stops, namely stops detected without a recorded reason.
Shop floor problem
Without structure, stops are noted at end of shift, sometimes end of week, from operator memory. Reasons stay vague: 'line problem', 'setting', 'other'. Micro-stops are not entered because too short to record by hand. Maintenance has no reliable history to prioritize interventions and stays reactive. Without a recorded reason, a stop cannot be analyzed. It feeds neither indicators nor action plans. The most costly losses stay hidden behind an average OEE that says nothing precise about what to fix.
Concrete example
Useful analysis goes beyond the OEE figure. It separates availability, performance and quality, then connects each component to observed facts: downtime, slowdowns, production and scrap. By comparing downtime reason duration and frequency over a period, the team can distinguish a one-off failure from a short recurring loss and prioritize shop floor analysis accordingly.
Operator role, maintenance role
Downtime qualification plays out at two complementary levels. On the operator side, entry should happen as close to the event as possible, on a shop floor tablet, with standardized reasons and sub-reasons preconfigured per line. The operator picks in seconds the category that matches what they observe. This entry captures the real context at the right time. On the maintenance side, the declaration can be enriched after intervention with a technical diagnosis (impacted component, actual duration, corrective action). This layer ties the stop to a machine history and feeds preventive action prioritization.
How a MES helps
A MES platform like IBRIXIO detects unplanned stops automatically from PLC signals (transitions running / stopped / fault). It offers fast reason entry on tablet with standardized categories and per-line sub-reasons. Unqualified stops are flagged and followed up. Over time, stops are cumulated by reason over duration and frequency, supporting objective prioritization of structural actions. The team moves from a reactive stance ('there was a breakdown') to factual prioritization ('this reason represents 18 minutes per shift, it is the costliest, we tackle it first').
Common mistakes
- Recording stops at end of shift from memory instead of capturing them at the source.
- Skipping micro-stops because they look too short.
- Using overly generic reasons ('other', 'line problem') that block any analysis.
- Mixing up detected and qualified: a stop with no reason has no operational value.
- Handling stops case by case instead of analyzing recurrence over several weeks.
Key takeaways
- Micro-stops often cost more than long stops.
- An unqualified reason is a hidden loss.
- Operator entry at the source captures the right context.
- Frequency × duration analysis reveals the real priorities.