OEE (Overall Equipment Effectiveness) is known in the industry as the standard metric for evaluating the efficiency of an asset or a production line. This indicator synthesizes three critical dimensions of operation: availability, performance, and quality. However, the conceptual simplicity of its mathematical formula often hides a deeper operational problem: most companies believe they are calculating OEE the right way, but in reality, they base their decisions on inaccurate, incomplete, and often inflated data.
The first obstacle to a correct OEE is not in the calculation itself, but in the low reliability of the data collected on the shop floor. In plants operating without an automated framework, information gathering depends almost entirely on the discipline and memory of operators, who record downtime events in shared network spreadsheets or basic MES software. This manual methodology generates delays and a distortion in the presented numbers.
The gravity of this scenario becomes more evident when industries transition from the manual model to an automated system connected directly to the machine controllers. Historically speaking, manually declared OEE values are typically 8 to 12 percentage points higher than the process reality. The moment automatic collection is activated, the indicator can suffer an immediate drop of 15 to 20 percentage points.
This abrupt drop does not reflect a real loss of efficiency, but rather the true operational reality. Faced with this challenge, industrial automation becomes a necessity in operations that require data integrity and auditing.
4 critical errors of manual OEE
The reliance on manual processes to measure industrial efficiency compromises the integrity of OEE through four common errors that frequently occur in factory routines:
Error 1: invisible micro-stoppages
Micro-stoppages are very short interruptions, usually lasting less than five minutes, caused by minor incidents that the operator themselves can resolve without the need for maintenance intervention. Examples include misaligned sensors, blocked conveyor belts, or minor material jams.
In manual logging systems, these stoppages are typically ignored. The operator usually lacks the available time to record events that last only a few seconds, creating an information gap.
However, the cumulative impact of these interruptions is significant. A three-minute micro-stoppage occurring every eighteen minutes throughout an eight-hour shift consumes 80 minutes of active production. This represents a silent loss of 16.7% of the available time. Without direct electronic monitoring of the machines’ signal outputs, these hours disappear from reports.
Furthermore, this cumulative effect compromises OEE in an unstructured way, as demonstrated in the mapping of these losses:
| OEE | Impact of micro-stoppages | Final indicator |
| Availability | The actual downtime is omitted from spreadsheets because the event does not reach the manual logging threshold (usually fixed at 5 or 10 minutes). | The availability pillar is inflated, hiding between 3% and 8% of the equipment’s actual idle time. |
| Unmatched Performance | Due to the lack of recorded downtime, the system assumes the equipment operated continuously but produced fewer parts. | The loss is diluted as a speed reduction, generating a distortion of 9% to 15% in actual performance. |
| Quality | The constant starts and stops of the production line alter the thermal and mechanical balance of the process. | The restarting of the equipment after each stoppage generates out-of-specification parts, leading to hidden losses of 2% to 5%. |
Error 2: logged “from memory”
In manual or digital systems lacking a direct connection to the machine, operators usually record the reasons and durations of stoppages retrospectively, often only at the end of the shift or hours after the occurrence.
This leads to simplifications and approximations. An event that lasted exactly 13 minutes and 42 seconds ends up recorded as “15 minutes,” while smaller interruptions of two or three minutes are frequently forgotten or grouped under generic justifications.
This lack of precision compromises data reliability, renders consistent analyses unfeasible, and weakens initiatives for operational improvements.
Error 3: hidden losses
The performance calculation of OEE compares the actual production cycle time with the ideal cycle time, meaning the theoretical maximum speed for which the machine was designed. A recurring error is the failure to update this ideal cycle parameter within the control systems.
As machinery ages or suffers wear and tear, it is common practice to lower the operational speed to prevent breakdowns or quality flaws. When this speed reduction is not accompanied by an update to the ideal cycle parameter in the OEE calculation system, the indicator begins to show a distorted reality.
In this scenario, a line operating at 85% of its theoretical capacity, even with no recorded interruptions, can falsely present a performance of 100%. The result is the hiding of a continuous loss of 15% in daily productivity, which goes completely unnoticed.
Error 4: lack of standardization
Without the automation of machine states, classifying the reasons for a stoppage becomes a subjective process. Faced with long and complex lists of engineering codes, operators from different shifts often classify the same event in different ways.
A mechanical jam caused by bearing wear, for example, might be logged as “Mechanical Breakdown” by one shift, “Operational Adjustment” by another, or simply as “Others.”
This inconsistency compromises the database’s reliability and prevents accurate analyses. As a consequence, investments and corrective maintenance actions can be directed toward the wrong root causes, reducing the effectiveness of decisions.
How automation transforms OEE into a strategic tool
The foundation of a reliable OEE lies within the PLC and field sensors. The controller processes physical input and output signals in real time, eliminating the dependency on manual logs to monitor machine states.
In this framework, monitoring occurs through internal logic that continuously analyzes the physical variables of the process:
- – Execution Signal (Running): Defined by the electrical activation of the main motors or by the detection of axis movement via sensors.
- – Blockage Signal (Blocked): Identified by photoelectric or inductive sensors active at the equipment’s discharge output, indicating that the downstream line has jammed and forced the asset to stop due to flow saturation.
- – Deprivation Signal (Starved): Mapped by level sensors that detect the absence of components or raw material at the equipment’s inlet, preventing the continuity of the production cycle.
- – Fault Signal (Faulted): Triggered by internal PLC alarms, door-interlock safety relays, or thermal trips of frequency inverters.
These signals are converted into instantaneous and reliable records. For example, optical counting sensors automatically record each unit produced, distinguishing between approved and rejected parts based on the activation of pneumatic reject pistons monitored by logical signals.
The integration among PLC, SCADA, MES, and OPC UA
For the signals collected by the PLC to aid in corporate decisions, the operational technology (OT) and information technology (IT) layers need to communicate in an integrated manner.
The SCADA supervisory system is responsible for real-time monitoring and data acquisition on the factory floor. However, on its own, it cannot obtain the necessary context to calculate OEE, since it does not identify which production orders are in execution or the ideal cycles for each SKU registered in the ERP.
This integration is performed by the MES system, which acts as the production’s intelligence layer. The standard responsible for ensuring this seamless and secure interoperability is the OPC UA protocol.
OPC UA eliminates dependency on proprietary drivers and standardizes communication between PLCs from different manufacturers and management systems. From a cybersecurity perspective, this protocol integrates end-to-end encryption, cryptographic message signatures, and strict access control via digital certificates.
In terms of operational ruggedness, OPC UA also offers data buffering capabilities. In the event of an industrial network connectivity failure, the gateway or local server stores production and stoppage logs at the edge and performs automatic synchronization with the MES as soon as communication is re-established. This prevents data loss and gaps in the OEE history.
High-precision time synchronization
The correct evaluation of micro-stoppages in ultra-high-speed lines requires exact network time coordination. The common NTP protocol, widely used in IT networks to keep clocks updated, operates with a precision of 1 to 10 milliseconds. While sufficient for administrative applications, NTP can experience severe variations and switching delays that make an accurate root-cause assessment of consecutive and simultaneous interruptions on the shop floor impossible.
A more advanced automation architecture adopts PTP (Precision Time Protocol) to timestamp data packets directly within the hardware circuitry of switches and controllers, achieving precision in the sub-microsecond range.
Managed switches with boundary clock support neutralize internal network latencies and ensure that all components, from the PLC to the MES database server, share the exact same temporal moment. In this way, the system identifies with microsecond precision which electrical fault or sensor signal triggered the chain-reaction stoppage of a line with multiple integrated machines.
What is needed to calculate a reliable OEE
Obtaining consistent data represents only half the journey toward an effective OEE. The other half lies in defining processes and a stoppage-cause structure that can be used quickly and clearly by operators.
To achieve this, it is necessary to develop a simple structure that can support reliable analyses with hierarchical levels. This organization facilitates daily use in operation and reduces classification errors:
- – Level 1 (Primary Category): Focused on executive-level overviews and macro responsibility assignments. It is restricted to a list of 6 to 8 options covering major domains, such as: equipment breakdown, batch changeover, lack of material, quality stoppage, planned maintenance stoppage, or operational problems.
- – Level 2 (System/Subcategory): Identifies the mechanical assembly or equipment area responsible for the stoppage event. Examples: equipment breakdown, pneumatic system, hydraulic unit, conveyor assembly, or sensor system.
- – Level 3 (Specific Root Cause): Provides the detailed breakdown that enables the application of structured corrective actions. Examples: sensor system, cable damaged by tension, optical misalignment, or hardware failure.
As you can see, the OEE calculation only fulfills its role as a continuous productivity diagnostic tool when it is separated from subjective estimates and manual records. Relying on these logs compromises the calculation’s availability, performance, and quality, creating distorted scenarios that hide inefficiencies with a direct financial impact.
To restore data integrity and elevate the level of operational reliability, OEE becomes an increasingly strategic KPI for the industry. Through it, companies can identify losses, guide corrective actions, and implement improvements capable of boosting productivity and overall operational efficiency.
