• Written by Muhammed Abdulla NC, CEO

A common objection to using AI in maintenance is that it is too difficult to measure its effectiveness, but this objection is misguided because it misunderstands the nature of maintenance work. In reality, the challenge lies not in finding metrics, but in choosing the right ones that move for the right reasons. A shift supervisor checking the maintenance log, for example, needs to know whether the planned versus reactive ratio is improving, and whether the mean time between failures is increasing.

Leading measures

Leading measures such as planned versus reactive ratio and mean time between failures are essential in maintenance because they indicate whether the maintenance team is being proactive or reactive. A high planned versus reactive ratio, for instance, suggests that the team is taking a proactive approach to maintenance, which can help prevent equipment failures and reduce downtime. The category manager comparing two suppliers, for instance, may want to consider their maintenance strategies and how they impact these leading measures.

Mean time between failures is another critical metric that can help maintenance teams identify areas for improvement. By tracking this metric, teams can identify patterns and trends that may indicate a need for more frequent maintenance or repairs. This can help prevent unexpected failures and reduce the overall cost of maintenance.

Why the headline metric moves for unrelated reasons

Most people get wrong the idea that the headline metric, such as overall equipment effectiveness, is a reliable indicator of maintenance performance. However, this metric can be influenced by a variety of factors unrelated to maintenance, such as changes in production schedules or raw material quality. As a result, it is essential to look beyond the headline metric and focus on leading measures that are more directly related to maintenance performance.

A closer examination of the data, for instance, may reveal that an increase in overall equipment effectiveness is due to changes in production schedules rather than improvements in maintenance. This highlights the need to carefully consider the context and potential biases when interpreting metrics.

Baselining before rollout

Baselining before rollout is critical to establishing a meaningful benchmark for AI metrics in maintenance. This involves collecting data on current maintenance performance before implementing any new AI-powered tools or systems. By doing so, maintenance teams can establish a clear understanding of their current performance and identify areas for improvement. The production team, for example, may want to review their current maintenance schedules and procedures before implementing a new AI-powered maintenance system, which can be accessed through our production page.

Baselining also helps to ensure that any subsequent improvements can be attributed to the AI-powered tools or systems, rather than other factors. This provides a clear understanding of the impact of AI on maintenance performance and helps to build a business case for further investment.

What to record when nothing was found

When nothing is found during a maintenance inspection, it can be tempting to simply record that the inspection was completed and move on. However, this approach misses an opportunity to gather valuable data that can help improve maintenance performance over time. By recording the details of what was inspected and what was found, maintenance teams can build a more complete picture of equipment condition and identify potential issues before they become major problems.

For instance, a maintenance technician may record that a particular piece of equipment was inspected and found to be in good working order, but also note that some minor wear and tear was observed. This information can be used to inform future maintenance schedules and help prevent unexpected failures.

What this does not do

This approach to measuring AI in maintenance does not provide a magic bullet for solving all maintenance problems. It is simply a way to focus on the metrics that matter most and ensure that AI-powered tools and systems are having a positive impact on maintenance performance. It is also not a substitute for human judgement and expertise, but rather a way to augment and support the work of maintenance teams. For more information on how to get started with AI in maintenance, please visit our contact us page.

Ultimately, the goal of measuring AI in maintenance is to identify areas for improvement and inform data-driven decisions that can help reduce downtime, improve equipment effectiveness, and lower maintenance costs. By focusing on leading measures and taking a thoughtful approach to metrics, maintenance teams can unlock the full potential of AI and achieve better outcomes.