Machine Monitoring & OEE Glossary: 30 Essential Terms
Machine monitoring is the automated collection of real-time data from CNC machines — run state, cycle counts, spindle load, alarms and downtime reasons — to measure productivity and equipment health. This glossary defines 30 core terms used in CNC monitoring and OEE programs, written in plain language so production managers and operators can align on what each metric actually means.
What are the core OEE metrics?
OEE and its three factors are the foundation of any machine monitoring program. The terms below define how effectiveness is calculated and the loss categories each factor captures. Read them top to bottom, then jump to connectivity, maintenance and shop-floor terms further down.
OEE
Overall Equipment Effectiveness is the headline metric: Availability × Performance × Quality, expressed as a single percentage of planned production time. It captures how much good output a machine produces versus its theoretical maximum. See OEE tracking for live calculation.
Availability
The share of planned production time a machine is actually running. It is reduced by unplanned stops (breakdowns, material shortages) and changeovers. Formula: run time divided by planned production time. Availability isolates downtime losses from speed and quality losses.
Performance
How fast a machine runs versus its ideal cycle time during the time it was available. It captures speed losses — reduced feeds, micro-stops and minor stoppages. Formula: (ideal cycle time × total count) divided by run time.
Quality
The proportion of parts produced that meet specification on the first pass, with no rework or scrap. Formula: good count divided by total count. Quality losses include scrap, rework and startup rejects after a changeover or tool change.
TEEP
Total Effective Equipment Performance measures OEE against all calendar hours — 24 hours a day, every day — rather than only planned production time. It adds a Utilization factor on top of OEE, revealing capacity hidden in unscheduled shifts, weekends and idle time.
OOE
Overall Operations Effectiveness is an OEE variant that uses operating time (including planned downtime) instead of planned production time in the Availability term. It gives a broader view of how operations use the time the plant is open, sitting between OEE and TEEP in scope.
What downtime and timing terms should you know?
These terms describe when and why a machine is not producing good parts. Distinguishing planned from unplanned downtime, and cycle time from takt time, is essential for accurate OEE and realistic capacity planning.
Downtime
Any period a machine is not producing parts when it was expected to. Downtime is split into planned and unplanned categories and is the primary driver of low Availability. Accurate downtime reason codes are what turn raw monitoring data into actionable improvement.
Planned vs Unplanned Downtime
Planned downtime is scheduled and expected: maintenance, changeovers, meetings, breaks. Unplanned downtime is unexpected: breakdowns, tool failures, material shortages, jams. OEE excludes planned downtime from its time base; reducing unplanned downtime is usually the fastest path to higher Availability.
Cycle Time
The actual time to produce one part, from the start of one cycle to the start of the next. Ideal (or theoretical) cycle time is the fastest sustainable rate. The gap between actual and ideal cycle time is the basis of the Performance factor.
Takt Time
The rate at which parts must be produced to meet customer demand, calculated as available production time divided by customer demand quantity. Unlike cycle time, takt time is set by demand, not the machine. If cycle time exceeds takt time, the line cannot keep up.
Micro-stop
A brief stoppage, typically under a few minutes, such as a chip jam, sensor trip or part hang-up. Micro-stops are usually too short to be logged manually, yet collectively they are a major hidden drain on the Performance factor of OEE.
Which connectivity and data standards matter?
CNC machines speak many protocols. These terms cover the open standards and native controller interfaces used to extract data, plus the German shop-floor data-collection acronyms you will encounter in monitoring tools.
MTConnect
An open, royalty-free standard that publishes machine data over HTTP in a common XML format. It is widely supported and vendor-neutral, but often exposes a limited, surface-level set of signals compared with a controller's full native interface.
OPC-UA
OPC Unified Architecture is a platform-independent industrial communication standard for secure, structured machine-to-machine data exchange. Common on Siemens and many PLC-based systems, it supports rich data models and is a frequent backbone for IIoT and MES integrations.
FOCAS
Fanuc Open CNC API Specification is the native programming interface for Fanuc controls. It exposes detailed, controller-level data — spindle load, program names, alarms, part counts — beyond what generic standards surface. CNC monitoring uses native interfaces like this where available.
Modbus
A simple, long-established serial and TCP/IP protocol used to read and write registers on PLCs, sensors and auxiliary equipment. In monitoring it is often used for peripheral devices — meters, I/O modules, environmental sensors — rather than the CNC control itself.
MDE
Maschinendatenerfassung (machine data acquisition) — the German term for automatically capturing machine signals such as run state, counts and alarms directly from equipment. It is the automated, sensor-and-controller side of shop-floor data collection.
BDE
Betriebsdatenerfassung (operational data acquisition) — capturing operational and labor data such as order status, operator actions, downtime reasons and part confirmations, usually via an operator terminal. BDE complements MDE by adding the human and order context machines cannot report.
What maintenance and condition terms should you track?
Monitoring is not only about output — it is about machine health. These terms separate two often-confused maintenance strategies and define the signals used to watch a machine's condition in real time.
Condition-Based Maintenance
A strategy that triggers maintenance when a measured condition — spindle load, temperature, vibration, runtime hours — crosses a defined threshold, rather than on a fixed calendar. xynLog supports this through condition-based alerts that flag abnormal signals as they happen.
Predictive Maintenance
An advanced strategy that forecasts when a component is likely to fail, using historical data and models to schedule service before breakdown. It differs from condition-based maintenance, which reacts to current thresholds rather than predicting a future failure date. See predictive maintenance.
Spindle Load
The percentage of a spindle motor's rated capacity in use during machining. It is a key health and process signal: rising load can indicate tool wear, material variation or fixturing issues, while sustained high load may stress the spindle and accelerate maintenance needs.
Utilization
The share of available time a machine is actively used, regardless of speed or quality. In TEEP, Utilization is the factor comparing scheduled production time to all calendar time. Used loosely on the shop floor, it often just means how busy a machine is.
What MES, ERP and Industry 4.0 terms appear in monitoring?
Machine monitoring rarely stands alone — it feeds and draws from broader plant systems. These terms cover the software layers above the machine and the connected-factory concepts that frame modern monitoring.
MES
A Manufacturing Execution System manages and tracks production on the shop floor — work orders, scheduling, traceability and performance — bridging machine-level signals and business systems. Monitoring data often feeds an MES, or a monitoring platform fills part of the MES role for smaller shops.
ERP
Enterprise Resource Planning software manages business-wide processes: orders, inventory, purchasing, finance and planning. Monitoring connects to ERP so real-time machine status and counts inform scheduling and costing. See ERP integration for how shop-floor data reaches business systems.
IIoT
The Industrial Internet of Things — networked sensors, controllers and machines that collect and exchange data to improve visibility and decision-making in manufacturing. CNC monitoring is a core IIoT use case, turning previously isolated machines into a connected, queryable data source.
Digital Twin
A live virtual representation of a physical machine, cell or process, continuously updated with real data. In monitoring contexts a digital twin can mirror machine state and history for analysis, simulation and what-if planning without touching the running equipment.
Edge Device
Hardware placed near machines that collects, processes and sometimes buffers data locally before sending it onward. Edge processing reduces latency and bandwidth and keeps sensitive data on-site — relevant for shops that prefer EU-hosted or fully on-premise deployments.
What shop-floor and signaling terms round it out?
Finally, a few everyday terms used on the production floor and in monitoring dashboards. They describe where the work happens and how machines and operators signal that something needs attention.
Andon
A visual signaling system — lights, screens or stack lamps — that shows machine and line status at a glance and calls for help when a problem occurs. Modern monitoring tools provide digital Andon dashboards that replace or augment physical signal towers.
Shop-floor
The physical production area where machines operate and parts are made, as opposed to the office or planning layer. The term often distinguishes operational, real-time activity ("shop-floor data") from aggregated business reporting in ERP or MES systems.
Operator Terminal
A shop-floor screen where operators view machine status, confirm production, enter downtime reasons and acknowledge alerts. It is the main interface for capturing the human context (BDE) that automated signals miss. See the operator terminal.
Machine Connector
The software interface that links a specific controller or protocol to a monitoring platform, translating raw machine signals into structured data. Native, controller-level connectors capture more depth than generic standards. See available machine connectors.
How does xynLog fit this glossary?
xynLog is an EU-hosted CNC monitoring and OEE platform that reads native, controller-level data — not just MTConnect surface signals — across common controls like Fanuc, Siemens, Heidenhain, Haas and Mazak. It pairs real-time OEE tracking with condition-based alerts and a plain-language AI assistant that can run on a cloud LLM or fully on-premise via Ollama.
Want to see these metrics live on your machines? See xynLog on your own CNC brand — book a demo.