Digital technologies in maintenance

Published:  14 August, 2023

The latest Talking Industry (TI) panel discussion discussed how digital technologies can be used in maintenance activities. What can you learn from machines using digital techniques, such as machine learning and artificial intelligence (AI)? TI Chair, Andy Pye reports.

Many companies are still not making the most of the opportunity to provide insight into machines by monitoring critical components that are subject to wear and tear.

Richard Jeffers got proceedings underway by explaining that one of his favourite topics is the digitisation of maintenance. Maintenance is the process of keeping plants and equipment in good working condition, so that efficiency is retained and/or the life of that asset is increased. Maintenance, he argued is not actually that complicated. But it is really hard work to get it right every day consistently.

The first generation was very much reactive, time-based maintenance. There was really no underpinning understanding of reliability theory of why components failed. And then in the 50s, 60s and early 70s. we moved into the second generation of maintenance. Driven by civil aerospace, there was an emerging understanding of reliability theory and a growth of preventive maintenance. This led to an increased asset reliability, lower cost of ownership and - particularly important in the civil aerospace - a reduction in aeroplanes falling out of the sky! Then followed the third generation of maintenance: a real understanding of reliability theory and how and why components fail, which led to the growing use of condition-based and predictive maintenance. Now, Maintenance 4.0 is the fourth generation of maintenance, the digitalisation of this process.

I asked why, if the basic theory goes back to 1968, why has it taken so long to put into practice? What do we need to do to get these techniques more widely adopted?

Richard Jeffers answered that the falling cost of technology makes it easier to deploy in lower criticality environments. What we see now is technology that was traditionally only available in in the aerospace industry or high value fleet operations, now being able to be moved into broader manufacturing environments. What really underpins the whole world of condition monitoring is that understanding of how components fail. Only 10 to 25% of components fail for time-related use - the other 75 to 90% fail for random reasons. And so condition monitoring and predictive maintenance is all about identifying and measuring a leading indicator of that random failure in a cost-effective and reliable way.

David Roddis emphasised that an obvious factor is education. Business leaders really need to start to understand the benefits of the digital age. If you are not bringing and using digital tools, you will start to fall behind your competition.

John Erkoyuncu concurred that skills development is key, looking at this in a strategic way from leadership to mid-management level to those may be coming into the organisations. How do we develop the ability to make decisions from a “digital 4.0 perspective”? Something that we need to focus on is the continuous transformation of businesses, going beyond the ad hoc introduction of technologies, and developing a culture where you are embracing new technologies coming into the organisation. I asked Erkoyuncu how, in training young engineers, he keeps up to date with fast-moving technological developments. How does he keep students up to date? And conversely, how can you influence the culture in a in senior managerial environments?

Erkoyuncu replied that Cranfield is a wholly postgraduate University, very much focused on working closely with industry. As an example, there is an MSc course called digital and technology solutions, which was developed with over two thousand companies that actually shared the kinds of skills requirements that they have.

* Talking Industry is sponsored by the Drives & Controls Exhibition, the #1 event for automation, power transmission & motion control. Taking place 4-6 June 2024, at the NEC, Birmingham. In association with Manufacturing & Engineering Week 2024. Drive The Future.


John Erkoyuncu, Professor of Digital Engineering and Head of the Centre for Digital Engineering and Manufacturing at Cranfield University

Richard Jeffers, Founder & Managing Director, RS Industrial

Dave Roddis, Senior Advisor - Digital Transformation, The MTC


Reliability theory and the digital mindset

Pye: “One of the perhaps unwarranted criticisms of academia in fast moving technology areas is how they keep up with developments in technology. And then you have the issue of standards lagging behind. These difficulties are in the background, impeding the implementation of something as new as industry 4.0.”

Erkoyuncu: “It is hugely challenging. And I think the key thing here is that we continue to communicate, to ensure that we are not stuck in what’s available today. We need to solve today’s challenges, but we also need to be a step ahead.”

Roddis: “Although we are talking about maintenance, when you are capturing data, you want to think about how to use that data throughout the business, not maintenance of the machinery, but utilisation as well. We call this the ‘digital mindset’”.

Jeffers: “For me, there is still an over-reliance on time-based maintenance in most factories that I visit, with people not understanding the failure modes that are impacting the assets, therefore using maintenance approaches that are actually going to induce early life failure rather than, you know, predict a random failure event. I agree completely with the points that John and David made around education around digital, but I think you’ve got to underpin that with education around reliability theory as well, because otherwise, you risk digitising something you don’t understand!” The conversation continued.... listen to the on-demand version of this Talking Industry episode to hear more the panelists’ comments, together with answers to the attendees’ questions.


Talking Points - the true cost of downtime

It’s often thought that you need to plaster your machine with sensors to get any useful information out of it, but it is often not as complicated as that: much essential data can all be accessed via a simple PLC.

To help maintenance teams work cost-effectively, it is important to have a view of an entire plant helping maintenance teams work easily and cost-effectively.

Richard Jeffers cited some work with the IMechE based on 700 returns from engineers within the UK. And we asked about the average cost of downtime and the amount of unplanned downtime that’s experienced so and that came back with the average unplanned downtime was 20 hours a week. And the average cost of downtime was £5000/hr. So that’s £100,000/week or £5 million/yr. If you are operating at a margin of 10% return on sales, you’ve got to make £50 million worth of incremental sales to recover the lost impact of that downtime.

To read the IMechE/RS ‘Industry in Motion’ survey please visit:


Sign up for the PWE newsletter

Latest issue

To view a digital copy of the latest issue of Plant & Works Engineering, click here.

View the past issue archive here.

To subscribe to the journal please click here.


"How is your manufacturing business preparing for a net Zero target?"