Every sector of industrial production is undergoing a real revolution in the digital and Industry 4.0 era. Maintenance is no exception.

Limiting production downtime is one of the most important aspects in enhancing the OEE (Overall Equipment Effectiveness) of a plant. New technologies provide the maintainer with a series of tools for enabling innovative approaches and excellent results.

Production downtime

Production downtime can be classified into two categories, foreseen and unforeseen.

The first category includes factory closures, as well as preventive and scheduled maintenance operations. The second includes all stops tied to various faults or problems, preventing a production objective achievement.

The latter is critical not only in terms of lost efficiency but for a series of related costs: emergency maintenance interventions, damaged company image and, in the event of non-delivery, possible customer loss.

The challenge for Industry 4.0 in the maintenance area is to guide operators in defining maintenance interventions based on the actual state of the factory and on data-driven forecasts.

Advantages of predictive analysis

Predictive analysis and maintenance are some of the Industry 4.0 responses to the production downtime problem. It does not stop there. The benefits of this methodology are multiple across various areas.

 

 Improve problem solving in the workplace: download our ebook "Ishikawa diagram  benefits"

 

Downtime reduction

The first, most evident advantage is the reduction of failure and, therefore, of unexpected downtime. This not only affects OEE but also savings on hard costs related to personnel and emergency maintenance materials.

Maintenance strategy optimization

Different maintenance policies have different criticalities: corrective maintenance is unforeseen and leads to high downtime while periodic maintenance can lead to maintenance that is too frequent or unnecessary.

Predictive maintenance enables planning interventions based on forecasting a machine's lifetime (RUL: Remaining Useful Life). This combines the advantages of the two previous methodologies: maintenance is carried out exactly when needed, neither too early nor too late.

Predicting and avoiding breakdowns also reduces stress taken on by the factory. This makes it possible to increase life expectancy and prevent a small glitch from reverberating, causing stress and breakage throughout the machinery (e.g., tool wear leading to excessive strain on the motor, causing it to break).

Concrete steps

First off, consider the equipment itself when implementing an effective predictive analysis.

In terms of 4.0, we are no longer talking about machinery but about CPS, or Cyber-Physical Systems.

This means that an object capable of gathering information within its own status and surrounding environment connects to other CPS networks to transmit and make such data accessible to the network.

The second point concerns data accessibility. The plant's informational potential should not be underestimated. Despite the large amount of data generated by sensors, equipment, and software, these are difficult for companies to access and even more difficult to view. The employees need to have a clear and complete view of the information to move forward in the analysis.

Predictive analysis data can be grouped into 4 categories:

  • Machine parameters: internal temperatures, consumed power, vibration intensity and frequency, oil level...
  • Plant data: temperature and humidity of the environment, use of each machine...
  • Historical data: frequency and fault types, history of machine parameters, performed maintenance operations...
  • Operating specifications: company knowledge on plant operation, desired specifications, minimum expected OEE...

Starting from this data, a dedicated software performs a predictive analysis to associate a specific failure probability and a forecast for future probability to a specific machine status.

Based on this forecast, maintenance is planned according to the company maintenance policy.

Conclusions

Reducing system faults lessens production downtime and saves on costs while increasing company efficiency.

Maintenance based on predictive analysis enables planned interventions for avoiding breakdowns without having to perform an excessive number of checks and repairs. Data collected for the analysis also provide useful insights for continuous improvement.

A final yet by no means negligible aspect involves the enhancement of plant safety. In fact, preventive intervention programming reduces the number of interventions carried out in complex or dangerous conditions. This results in lower accident probability.

 

New call-to-action