Aeronautical Maintenance Training EASA Part-66

Predictive maintenance consists of anticipating future failures in equipment, objects, systems, etc. In concrete terms, it involves anticipating a breakdown or malfunction thanks to the accumulation of a set of data.
To this end, the equipment is equipped with sensors connected to the internet and informing the user and/or the maintenance department in real time so that they can monitor the condition of the equipment and, if necessary, carry out repairs or replacements in due course.

Several advantages of predictive maintenance :

By analysing the data provided by the sensors, it is possible to anticipate a breakdown as soon as there are warning signs on the machine, enabling parts to be changed at the right time and reducing costs by changing parts unnecessarily.

It avoids, for example, the immobilisation of a broken down production system (with the resulting financial repercussions).
According to a McKinsey report, maintenance costs can be reduced by 10 to 40% and the number of breakdowns can be halved.

What data does predictive maintenance rely on?

The parameters monitored are, for example, the temperature, the pressure of fluids in pipes, the level of vibrations they emit or the sound of cracking in the material of parts or structures (micro-cracking for example) inherent in the life of the material. By opting for predictive maintenance, the company has the possibility to detect the defect of its machine before it occurs and to intervene before the breakdown occurs.

Predictive maintenance, artificial intelligence and machine learning

Artificial intelligence provides the analytical tools to detect anomalies in the sensor data that may indicate a failure. It is therefore essential to the proper functioning of predictive maintenance and is the source of cost reduction.

Indeed, maintenance departments can account for 60% of operational costs. It is therefore essential to reduce these costs in the context of sustainable development.

Machine learning is a branch of artificial intelligence that uses machine learning algorithms to analyse data and diagnose faults at an early stage.

What are the benefits of predictive maintenance?

Predictive maintenance, a source of ROI (return on investment), has several consequences:

•    the reduction of breakdowns and therefore of equipment downtime
•    extending their service life
•    improved equipment reliability and therefore production optimisation
•    better management of spare parts stocks
•    reduced maintenance costs.

Predictive maintenance in industry

Predictive maintenance is particularly dominant in the industrial market as it applies to rotating machinery. The coronavirus health crisis has also accelerated the deployment of predictive maintenance in industry, in order to avoid unnecessary travel by technicians.
Predictive maintenance is increasingly seen as a strategic objective. One in five companies (20%) has set up an in-house team to deploy predictive maintenance projects.