Predictions on Predictive Maintenance: it will fail… unless
This is the first article in a series on the hot topic of Predictive Maintenance, drawing on the practical experience of Siveco working with maintenance improvement in China and the Asia region in the past 20 years.
Signal vs. Noise: the fuss about predictive maintenance
There has been a tremendous fuss about Predictive maintenance this summer. Consulting firms and sensors suppliers seem to rediscover maintenance after so many years of staying carefully away from this complex – and often non-bankable – topic. Time has changed now, and thanks to the limitless power of AI and IoT ubiquity, the maintenance nut will soon be cracked (possibly by a cobot nutcracker).
Indeed the combination of new achievements in calculation power, connectivity and increases in availability of data science technology yield great possibilities for collecting and interpreting machine data and other technical information. But will predictive maintenance free us of all breakdowns, quality losses and safety risks?
Predictive maintenance – i.e. the ability to provide a reliable forecast for a failure or to alert the operator about a change of condition of its equipment – is nothing new. What we merely observe under the Industry 4.0 shift is simply an increase of its availability and a rebranding of familiar maintenance approaches such as condition-based maintenance.
In any case, there is no point in replacing ‘traditional maintenance’ with ‘predictive maintenance’, and believing all the problems will disappear doing so. Never mind, consultants and suppliers relentlessly describe Preventive maintenance overthrowing Corrective maintenance, and nowadays Preventive maintenance being replaced by Predictive maintenance. Funny moment at a conference this summer: the presenter announcing the replacement of Predictive maintenance by “Proactive Maintenance” which would enable “new business models”, without of course elaborating further on what these new business models will do – except being new. Empty narrative, boring story.
At the Smart Maintenance Conference 2019, Siveco VP Paul Wang insisted on maintenance methodology
Yet, there is no doubt that the scope and sophistication of maintenance strategies made available and affordable to industrial owners and operators is expanding – and this is good news. The question is, then, how to be correctly positioned to capture as much value as possible from the recent technological progress and identify & implement the most useful tech in one’s organization?
Digitalization: being right and being big
At another conference, a major international automation vendor described a case study on predictive maintenance (conveniently, the case study displayed no name, no data), where the pilot stage has been reached. Why were Siemens and its client not able to move past the ‘pilot purgatory’ asked the audience? Answer: “Because at that point, we realized the client’s processes were not digitalized enough”. A predictive maintenance pilot – with top-of-breed Siemens tech – must have been a very expensive way to realize just that. And we see a very familiar pattern emerging again: the ability for the supplier to blame the client for its lack of maturity, a convenient fallacy we’re denouncing regularly in this newsletter.
Siveco COO Guillaume Gimonet sharing on practical maintenance improvement experience at the Process Intelligent Manufacturing Summit 2019
This sets the stage for a fundamental truth: no technological shift will be possible without a minimum of digitalization already achieved. Think about a simple, yet well-established and centralized database of asset register, maintenance plans, maintenance workflow and key performance metrics. In many organizations, this information is still scattered across several media (paper, excel, CMMS,…) and several department. Digitalizing and centralizing the right work processes and the useful amount of information with the ad hoc structure are the true enablers of the Industry 4.0 shift.
And for this matter, ‘right data’ is better than ‘big data’. Indeed, contrary to the Internet industry, we cannot trick machines into giving away their ‘personal data’ for free… so each and every data has, practically, a cost. And how right data is often a matter of perspective, better framed through your maintenance plan.
This article will continue in the next issue, touching on information overload and the resulting “Cassandra effect”, looking at Smart O&M as a predictive maintenance enabler, and concluding with a discussion on the different types of predictive maintenance models.
Our Siveco experts are frequent speakers at industry events, conferences or workshops for specific clients on the topics of Smart solutions, IoT and Predictive Maintenance. Do not hesitate to contact us at info@sivecochina.com to discuss this subject!
