How to reduce downtime of machine tools by intelligent predictive maintenance?

Fagor Aotek participates in the 7th frame program European project nr. 314548 “Power consumption driven Reliability, Operation and Maintenance optimization”.

A number of companies from different countries collaborate in this project related to monitoring of machine components.

The main project objective is to reduce downtime of machine tools by intelligent predictive maintenance. In normal practice, maintenance is done at scheduled intervals or directly after detected malfunctions. In the first case, the schedule does not take into account the operational conditions of the machine and in the second, the failure can derive in unexpected effects on the worpieces or in the machine itself.

There is a trend on predictive maintenance towards smart strategies that minimize downtime. The Power-OM project proposes the use of electric current consumption monitoring and profiling to implement Condition based Maintenance (CbM).

This intelligent predictive maintenance strategy tries to assess the “health” of the machine through some short, standard test programs that combine diagnostics for the most important components to generate a report of the condition of the machine. During these tests, a number of variables coming from sensors, both internal to the CNC-Drives system and some external (accelerometers…), are sampled and stored.

After processing, the result of the test (a “fingerprint” of the machine) is generated and can be compared with some patterns to evaluate health figures related to the most critical components (mainly spindle and linear guides in machine tools).

Combined with this test, executed at periodic intervals (daily, weekly…), operational data should be sampled and stored. This provides information to correlate failures or malfunctions with operating conditions (power, speed ranges …) trying to identify the parameters of a process leading to potential malfunction .

Both the operational data stored as well as the fingerprints can be sent to a remote server where, based on knowledge from the same or similar machines, the CbM algorithms can suggest taking maintenance or repair procedures, scheduled in such a way that the production is not stopped.

To achieve these objectives, the FAGOR CNC8070 together with AXD/SPD digital drives are very well suited, and all the internal data needed can be sampled at high rates.

To access the external signals, taken by the ARTIS GEM-OA devices, a real time bus is necessary. The communication between Fagor Automation and Artis is done in CanOpen at cycle times between 4-12 milliseconds. Artis has developed the CanOpen slave protocol and data flows at the specified rates between the components. The information is exchanged in both directions and can be configured flexibly.

It should be noted that this communication framework establishes a good starting point to continue the collaboration between Fagor and Artis in the process monitoring area, where added value in the field can be integrated in Fagor CNCs.