<img height="1" width="1" style="display:none;" alt="" src="https://dc.ads.linkedin.com/collect/?pid=89122&amp;fmt=gif">

Get in touch

If you would like to know more about Certus, our solutions, services, or how we can help your business, please fill out the form below.

LATEST POST | 3 Jul 2020
AUTHOR | James Hartwright

The Four Methods to Improve Asset Intelligence and Predictive Maintenance Efforts

Post Featured Image

Learn about the four methods that enable leading organisation to improve asset intelligence and make the best predictive maintenance decisions.

Asset Maintenance is a balancing act. It is important to carry out regular preventive maintenance to ensure machines operate efficiently and to identify potential issues before they cause disruptions. However, doing too much preventive maintenance leads to inefficient use of recourses and can, in some cases, even cause breakdowns that would have otherwise not occurred.

Data suggests that 30% of preventive maintenance activities are carried out too frequently. However, the solution is NOT to go back to purely reactive maintenance processes that only address problems instead of looking to prevent them.

Instead, leading asset intensive businesses are increasingly embracing what is called predictive/prescriptive maintenance to find the right balance to their maintenance efforts. Leading-edge AI technology is helping them get more efficient at this every day.

Predictive maintenance is about using AI to leverage all the data we now have access to from our machines to determine the best maintenance frequency and processes based on the asset’s condition, anatomy and performance trends.

There are four methods that enable organisations to make the best predictive maintenance decisions by improving asset intelligence.

1. Data

For predictive maintenance to work efficiently, we first need to bring all the data together, into an asset-centric view. This includes information about the asset's behaviour, age, underlying content, maintenance history, location and more. The key is to not only capture this data but also to connect it and make it available to the people in the organisation so they can use it in their decision making. It also is essential to monitor the data and regularly review data quality.

2. Visualisations

Visualisation is all about making the data usable. By visualising data in charts, images and dashboards and putting it into context with other data points, people can make sense of the information and take action.

An important aspect of visualisation is to think about how it will be used. This includes, for example, to think about what kind of device the visuals will most likely be viewed on (complex graphs on a small screen are usually not a good idea).

3. Leverage Knowledge

Leveraging knowledge starts with gathering knowledge. This includes manuals and work order notes, but also the information in people’s heads. Almost every asset intensive business I've come across has at least a few people who have been on the job for several decades and have an incredible wealth of knowledge stored inside their heads – and many of them will be retiring soon. Therefore, leaders must start to think about how they will gather that knowledge to make sure it does not leave the business with the person.

However, it's not as simple as asking your experienced engineers to write down everything they know and store all manuals in a specific location. Instead, the knowledge needs to be documented in a format that is accessible and user-friendly for the people who will need it most so it can be leveraged.

4. Analytics

Analytics is about reviewing the data that has been gathered and extracting actionable insights. For this method, humans and AI work together to looks at historic outages and failures, look for patterns that might predict it, implement rules to alert if it happens again, check the validity of the alert, feed back and refine and build an analytical model for further refinement. 

A vital aspect of this is the collaboration between AI and humans. AI will only ever be as good as the data it was trained with and depends on further input to improve over time. Experienced humans can provide this input until the AI has reached a point where it can operate efficiently on its own.

These are the four fundamental methods of improving asset intelligence and enabling efficient predictive maintenance. However, I believe there is a fifth method that is incredibly important, not just for improving asset intelligence but for AI initiatives in general.

To find out what this fifth method is, and to learn more about the other four and asset intelligence in general, visit our Maximo in Action showroom page. On the page, you will find a video recording of my talk at the recent Maximo User Group conference as well as many other useful Maximo resources.

 

VIEW THE MAXIMO IN ACTION SHOWROOM



Tags