Predictive analytics for CNC machines

The Context

How to increase availability of CNC machines and avoid unexpected failures?

Our customer, a multinational manufacturer of automotive parts, operates over 100 CNC cutting and milling machines at their plant in Belgium. In order to keep up with the demand, ensuring machine availability and increasing productivity are key priorities.

Condition monitoring through predictive analytics could be an interesting way to improve maintenance planning and avoid unplanned downtime. This however poses a serious challenge as these machines come from various manufacturers, all with a different design and control system.


Download the summary of this case study in English / French  / Spanish

Diagnosis of increasing vibration on 2 points

Our Solution

Condition Monitoring on CNC machines is a complex matter due to the huge variety in machine designs and controls. To enable continuous online monitoring and avoid having to create dedicated interfacing protocols for each machine, P4A developed a standalone online monitoring system. This system can be installed on many different types of CNC machines and thereby also reduces the risk of receiving input that doesn’t meet the requirements for proper monitoring.

The data acquisition system installed consists of 10 vibration/temperature, 2 speed, 2 current and 2 voltage sensors all collecting real-time data directly available on P4A’s web-based asset data analytics platform.

After a first phase of data collection our team of data scientists built the first algorithms to monitor the key components of the machine and detect changing patterns. Alarms were set and trends could be followed.

A first diagnosis soon followed when the vibration levels on 2 measurement points started to increase indicating a potential problem at the right ball slide rail.

As our prognosis trends indicated a quite rapid evolution of the situation our intelligent system recommended inspecting the right ball slide rail.

Unique, online and standalone data acquisition system
including vibration, temperature, speed, current & voltage

The Benefits

Today, most CNC operators are fully dependent on the OEM for maintenance schedules and operating advice. Furthermore, delivery time of spare parts can be very long resulting in important downtime, loss of production, and additional costs to compensate these losses.


Thanks to our standalone monitoring system with automatic predictive analytics the asset manager is now able to make objective decisions based on real-time data and move to condition-based predictive maintenance.  


Furthermore this same asset data can also be used to monitor other performance deviations, wear of the cutting tools and upcoming problems allowing better asset management and more reliable operations at lower costs.

During the inspection the team discovered a pile of chips that could damage the rail

Contact Us

Contact us for more information at +32 (0) 81 71 99 82 or send us a message via the contact page.