Evolving From Predictive to Liquid Maintenance in Postmodern Industry

Introduction

Liquid Maintenance is the Predictive Maintenance (PdM) evolution to thrive in the incoming industrial era. Industry, and subsequently, maintenance and, namely, PdM are unready to face the near-future incoming challenges since they are anchored in an obsolete paradigm alien to the incoming cyber-physical reality and unfit for unbelievable data density. In addition to this, PdM is wormed by philosophical hiddenness around Time and taxonomies abuse; it is not the sound subdiscipline it appears to be. So PdM professionals are doomed to dialogue and get along with AIs if we want to break our human predictiveness ceiling glass and keep PdM improving. In this paper, the authors will explain not only the turn of the tide but how to flow towards a non-essentialist PdM paradigm.

Maintenance happens in a non-deterministic polynomial-time hard (Abedi et al., 2017). In practice, it points out that early and tardy maintenance works are the norm rather than the exception. More even, industry keeps stuck to the Adam Smith paradigm of labour division extended to maintenance and the way it’s understood. Furthermore, the taxonomy as used in industry so many times is too close to Aristotelian verbiage and essentialism which makes it a somewhat immature field of study. And, to cap it all, meagre effort has been put in determining the way from scientific research to taxonomies, which is a long time encysted issue (Fales, 1979). Standardising the whole of maintenance activities is not quite different from setting a quality standard and shares a weak point with it. Great ideas as Total Quality Management not resulted so great when implemented: rigid and strong quality standards give room for a group feeling judged by the standards of others, so, rather than spreading quality it raised dissent and distrust (Brown & Duguid, 2017, p. 135).

Traditionally, what made a factory to be a factory was to be a firm’s place where machines produce things -hopefully, in more efficient ways everyday. However, old-fashioned firms are disappearing and the whole industry is diving into a reality where the machines, the goods and foremost the places are belittled by the importance of their digital dimension, sustainable chains of value and the flexibility that smart-tech brings to business opportunities. By the same token, the importance of AI in PdM becomes paramount. Even if it’s needed to demystify the breakthrough technologies that will revolutionise everything, the machine learning use in PdM, as in 2022, has grown exponentially (Redmon, 2021; Carvalho et al., 2019). It seems not to be a critical juncture but a structural shift. According to the McKinsey Global Institute, more than 60% of all manufacturing activities in the late 2010s could have been automated with the automation technology then (McKinsey Global Institute, 2016; 2017a). So, it’s not surprising that the value of 4IR techs is expected to share $3.7 trillions in value and become the next economic growth engine (McKinsey Global Institute, 2017b; Leurent & De Boer, 2018).

Physical, digital, and even biological entities are getting blurrier while Time remains indefinite: in such scenarios PdM can’t help melting with the 4IR and become a Liquid Maintenance.