All Systems Operational 

Predictive vs Preventive Maintenance

Share on twitter
Share on linkedin
All industrial facilities carry out maintenance. Two common types are preventative and predictive, and knowing the difference between the two and understanding which is best for your organisation is key.

Preventative Maintenance

Preventative maintenance is work which is performed, usually at regular intervals, on equipment to keep them in check. Work is performed regardless of the status of equipment. The aim of preventative maintenance is to prevent breakdowns and identify any parts which need changing.


  • Parts are usually kept in good condition
  • Can help to avoid catastrophic breakdowns by performing regular upkeep

Although preventative maintenance can help reduce equipment downtime by identifying issues early, there are some disadvantages with this approach too…


  • Checking every piece of equipment requires a significant amount of time and money.
  • Work is performed when it is not needed.

Predictive Maintenance

Predictive Maintenance on the other hand, involves the process of identifying issues before they occur, and performing work as and when it is needed. Predictive maintenance is a data-driven approach, requiring information that describes the condition of the asset to determine what is ‘normal’ operation. Therefore, unusual performance or operation can be flagged and investigations can begin. The aim of predictive maintenance is to detect issues before they cause a breakdown, reducing the downtime of an asset.


  • Work is targeted as and when it is needed.
  • Less downtime due to work being performed only when needed.


  • Requires data collection tools in place.
  • Requires good quality data.

Predictive vs Preventative Maintenance

We can all agree that both predictive and preventative maintenance provide a better approach than reactive. Reactive maintenance involves fixing an asset after it has caused issues and breakdowns. It’s not an ideal way to maintain assets. Reactive approaches increase asset downtime, and jobs are usually larger and more catastrophic.

Preventative maintenance is usually cheaper to set up due to the data collection requirements of a predictive approach.  However, over time, time and labour costs of preventative maintenance are likely to outweigh the set-up costs of predictive approaches.

Not only is the predictive approach a maintenance tool, but can also educate workers on the behaviour, operation, and expectation of systems. By increasing understanding of systems, as well as how systems in the workflow are interlinked, one is more informed when implementing control and management strategies.

Once set up, predictive maintenance can be used to provide detailed actionable insights, automated alerts and maintenance guidelines.

How To Implement Predictive Maintenance

A lot of buzz surrounds the phrase ‘Predictive Maintenance’, however, before jumping on board, it’s essential to understand the requirements and have a structured plan in place in order to fully reap the benefits.

Firstly, identifying assets that will benefit from predictive maintenance. Starting off with a few assets which have high maintenance costs, play a significant role in the production line, or are prone to breakdowns can provide a starting point.

An accurate predictive maintenance system will require good quality data. Establishing a data collection plan, how and what will be collected, what hardware is required, along with how data will be stored is critical.

This data can be analysed by using methods such as Failure Modes and Effects Analysis (FMEA). FMEA focuses on identifying failure modes and related risk of failure. This method looks at failure severity and occurrence, along with how failures occur and impact of potential failure. Not only does this allow for failures to be predicted, but also allows for understanding of corresponding risk, allowing prioritisation of key assets.

Predictive models built can improve over time, as more data around the assets performance becomes available. This allows for more behavioural characteristics, error codes, failure modes and more insight into the asset’s daily operation.

If you’re interested in how you can build robust data collection systems, and start your journey to a fully predictive maintenance system, contact

Victoria Mawson
Victoria Mawson
Share on twitter
Share on linkedin

Further Reading

Would you like to find out more about the Hark Platform?

Subscribe to Our Newsletter

Stay up to date with the latest industry news, platform developments and more.