Articles
Know-how on spindle monitoring for CNC machining centres.
Vibration diagnostics, control integration and economics, written by the software architect of our system. No marketing promises, with formulas and sources where they exist.
Our system
From vibration to alert: our system's processing chain according to ISO 13374
Between the raw signal on the spindle housing and the alert in the dashboard there are five clearly separated steps. Each one is traceable, and every threshold is stated openly.
The edge box for condition monitoring: design, tasks and behaviour during network outages
The edge box on the machine reads the CNC control, measures vibration, calculates features and sends only those. If the network fails, it buffers for at least seven days and then sends everything without gaps or duplicates.
Alerts with an explanation: what our dashboard shows and what the maintenance engineer does with it
A red light on its own helps nobody in maintenance. Our alerts say what stands out, where, by how much, how urgent it is and what to inspect. And they ask back whether they were right.
Learning the baseline: what happens during the learning phase and what to watch out for
Every spindle vibrates differently. That is why our system first learns how this one spindle vibrates when healthy, and only then starts assessing. Here is what that means in practice.
Mixed machine fleet: Siemens and HEIDENHAIN, old and new, in one view
Hardly any plant has only one type of CNC control. We show how our system brings SINUMERIK and HEIDENHAIN together through a common data contract, and what is possible on older machines without OPC UA.
Condition monitoring pilot project: how it works from the inside
A pilot project is not a software download. This article shows step by step what happens before, during and after the visit to your machine, what we rehearse beforehand and how a pilot ends cleanly.
Spindle and bearings
Spindle monitoring on CNC machining centres: the guide
What spindle monitoring can detect, which measurement equipment it needs, how vibration becomes an alert with a stated reason, and where the limits are. The overview, with links to every in-depth article.
Detecting rolling bearing damage early: what envelope analysis and bearing defect frequencies reveal
Incipient bearing damage produces tiny impacts at a fixed repetition rate. If you know which frequency to look for, you will find them in the envelope spectrum long before the bearing becomes audible.
CNC spindle making noise: possible causes and what you can check yourself
Not every new spindle noise is bearing damage. A systematic check narrows down the cause. And by the time the bearing becomes audible, the damage is usually already advanced.
Order analysis on the spindle: unbalance and bearing frequencies as multiples of running speed
A spindle rarely runs at the same speed for long. Order analysis therefore measures vibration in multiples of the shaft frequency. That way unbalance is always order 1, whether the spindle is turning at 3,000 or 12,000 rpm.
Method
Fixed thresholds, 3 sigma or a learned baseline: how alarms in condition monitoring are triggered
An alarm is only as good as the rule that triggers it. We compare fixed limits, the 3-sigma rule and a robustly learned baseline, and show where machine learning helps and where it does not.
Vibration analysis on machine tools: what ISO limits can do and where they stop
The ISO vibration standards provide a framework for the overall condition of a machine. Early diagnosis of spindle bearings needs more: trending against the machine's own learned baseline and frequency-based features.
Condition monitoring vs predictive maintenance: what the difference is and what a prognosis can deliver
Condition monitoring tells you how the machine is doing now. Predictive maintenance wants to tell you how it will be doing later. We define the terms and are honest about where prognosis reaches its limits.
RMS, crest factor and kurtosis: vibration features explained clearly
Every vibration feature answers a different question, and every one has a blind spot. If you know when a feature fails, you can combine them correctly and assess them against the learned baseline of the machine.
FFT in vibration analysis: sampling rate, resolution and window function for practitioners
A spectrum is only as good as the decisions made before it. Sampling rate, window length and window function determine whether two closely spaced bearing frequencies appear as two lines or as one blurred hump.
Integration and IT
Reading SINUMERIK and HEIDENHAIN via OPC UA: prerequisites, options, security
Which CNC control needs which option, what a read-only user looks like and why OPC UA is not enough for bearing diagnostics. A guide for maintenance and IT teams, based on the manufacturers' manuals.
OPC UA vs MQTT: which protocol belongs where for production data
OPC UA and MQTT solve different problems. OPC UA describes and reads the machine on the shop floor, MQTT carries condensed results securely and outbound-only to the platform. The computing happens in between, on the edge box.
Security and compliance
Read-only: how we technically rule out write access to the CNC control
A monitoring system on the CNC control must not be able to change anything, not even through a software bug. That is why we enforce write protection in four mutually independent layers.
Device certificates and signed updates: how the edge box identifies and protects itself
Every box has its own certificate whose key never leaves the box. It only accepts updates if the signature is valid, the version is newer and the package arrives unaltered.
Cyber Resilience Act and IEC 62443: what machine builders and operators of condition monitoring should know
The Cyber Resilience Act makes manufacturers of networked products responsible for their security. As an operator you should know which questions to ask a condition monitoring supplier and which answers you can expect.
Economics
What does an hour of machine downtime cost? How to work it out for your plant
The machine hour rate does not answer the question of downtime costs. To derive the cost of an hour of downtime properly, separate time-dependent from event-dependent costs and work with your own figures.
OEE on CNC machines: where availability is really lost
OEE is only as good as its definition. If you separate availability, performance and quality cleanly, you can see which losses on CNC machines really count and where spindle monitoring helps.
Other assemblies
Feed axes on CNC machining centres: wear patterns and their signals
Wear on feed axes first shows up as a loss of accuracy, not as a failure. Backlash and friction are the two quantities worth understanding and comparing over time.
Tool changers on CNC machining centres: failure patterns and early signs
A tool changer often announces problems in advance: through longer or fluctuating change times, isolated fault messages and changed drive load. If you collect these signs, you usually find the cause before the machine stops.
Coolant monitoring on CNC milling centres: metalworking fluid and spindle cooling
On a milling centre, the word coolant covers two different circuits. The metalworking fluid needs care and regular measurement, while spindle cooling determines temperature stability and dimensional accuracy.
Next step