Dr. Ahmet Yalçınkaya
Molino Mechanical Industry and Trade Inc.
“Industry 4.0 promotes Smart Predictive Maintenance (SPdM) and smart manufacturing as machines are connected as a collaborative community in smart factories generating a great potential for Predictive Maintenance (PdM). Transition to Industry 4.0 is a dynamic process and continues. PdM plays a fundamental role in that transition, benefitting also from the developments of that transition in terms of data collection and data analysis.”
Total Productive Maintenance and its Predictive Maintenance tool
Maintenance is defined in many ways depending upon which parameter one is focused, but in general, it can be defined as “all kinds of activities such as repair, replacement, inspections and similar to maintain the functionality of equipment and buildings throughout their expected lives”.1
These activities can be carried out using different methods which differ according either the conditions faced or the strategy selected. If we classify them, we can see that although a complex structure is observed, they can be grouped under three main methods applied2. A classification of maintenance activities is shown in Figure 1, and a summary given below.
Breakdown maintenance which is also called “reactive maintenance” is an emergency maintenance, is corrective and consequently unscheduled. It is carried out whenever a breakdown is faced. The problem is solved either in repairing or replacing the problematic element.
Planned maintenance is scheduled, and depending on the strategy, technology and present situation reports, one of the three planned methods may apply. One is “Protective maintenance (PM)” which is also called “preventive maintenance (PM) ,and “proactive maintenance (PM)” to point out its precautionary character, or “periodical maintenance (PM)” to emphasize that it is scheduled having daily, weekly, monthly, yearly maintenance periods. “Predictive maintenance (PdM)” making use of measurements and monitoring to predict the eventual breakdown times for taking precautions before such breakdowns occur is the other strategy which may be decided for. “Revision” is a case when maintenance is not sufficient anymore to maintain the functionality and extend the life of equipment and parts.
Redundancy is not a direct maintenance method, but a supporting activity to make maintenance and replacement possible.
Predictive maintenance (PdM) is a new approach and has in application briefly 3 phases , :
• Measurement phase, where critical points for measurement are identified, measurements are performed using NDT methods like oil analysis, temperature analysis through infrared thermography, ultrasonic test, vibration analysis, noise analysis, measurement of pressure and current differences, and also visual inspection.
• Analysis, where results of the measurements are analysed and failure source is determined.
• Repair, where fault is evaluated and necessary repair is carried out. This phase should also include a “planning” sub-phase.
Figure 1: Classification of Maintenance ActivitiesSource: Ahmet Yalcinkaya, Etkin Bakım Sistemi, Konya, 2013 (from Ilhan Or, Total Productive Maintenance Management)
A concept not mentioned among maintenance methods above is “autonomous maintenance”. It is not a direct method comparable to the described ones, but is an application to emphasize the degree of operational sufficiency. In autonomous maintenance, the equipment operator carries out some of the maintenance activities by himself/herself.
For planned maintenance, a strategy, a schedule and an execution program and method is inevitable. Planned maintenance cannot be performed like breakdown, and needs a sophisticated maintenance management strategy to guide, control and evaluate maintenance activities. Companies generally try to establish an “Efficient Maintenance System” to be able to manage maintenance.
Elements of an efficient maintenance system would be maintenance policy, materials control, work order system, equipment records, maintenance (both scheduled and unscheduled), work planning and scheduling, control of incomplete tasks and prior system, evaluation of maintenance in order to achieve maintenance goals such as minimizing time loss, maintenance cost, stopping time and wearing of equipment3.
The concept of Total Productive Maintenance (TPM) presents a proved, efficient, and well defined philosophy, methodology and technique to establish such a system. It includes technically the planning, monitoring, evaluating, management and development elements of an efficient system.
TPM needs well definitions, continuous and reliable information (and data depending upon the maintenance type) and analysis, both technical and managerial, to be able to yield effective results.
So, a master plan is necessary for establishing TPM; a plan including independent improvements, autonomous maintenance, planned maintenance, education and training, quality maintenance, new equipment management, and TPM plan for offices2.
The nature of TPM necessitates a blended methodology for maintenance to gain optimum benefit considering plant conditions, production plan, manpower, finance and similar parameters. Despite this, PdM is the method to support TPM philosophy at most due its analytic character, and is highly preferable for plants where continuous production is urgent as other methods will be insufficient in such cases.
TPM has many benefits for a plant, but the main parameter measured and calculated to ensure the efficiency of the system is the Overall Equipment Efficiency (OEE), which is formulated based on various defined periods, durations and times in usually minutes so that records of these durations should be kept very well.
Industry 4.0 and maintenance shortly
Industry 4.0 known as the fourth generation of industrial activity “represents the fourth industrial revolution on the way to combine modern industries with Cyber Physical Systems (CBS) , Internet of Things (IoT) and Internet of Services (IoS)”, and is characterized by smart systems and Internet based solutions. In an Industry 4.0 factory, machines are connected as a “collaborating community to collect, exchange and analyse data”4.
If we remember the historical process of industrial revolutions and try to establish a correlation between them and the mainly applied or suitable maintenance method corresponding to it, we can summarize the development as given in Figure 25.
Figure 2: Industrial Revolution TimelineSource: www.researchgate.net/publication/322369285
Facilitating effect of Predictive Maintenance on Industry 4.0
Returning to Predictive Maintenance (PdM), we can see that its application can be also defined based on the data structure or the data collection and analysis method it uses.
Predictive Maintenance (PdM) may be “rule based” which is actually “condition monitoring” where sensors collect data continuously and send alerts according to defined rules and thresholds. This would not meet the expectation if a factory or plant is desired to be smart.
On the other hand, predictive maintenance may be “machine learning based” which relies on large sets of historical data or test data combined with special machine learning algorithms6. It is possible in this way to run different scenarios and predict what will go wrong when, and generate alerts to point it out.
A Machine Learning Process consists of the successive data collection, feature extraction and reduction, model creation, model validation, and deployment steps  where the main goal is to run the created model and improve the system in relaying back discrepant behaviours. So, the loop is improving the behaviours instead of just repeating them.
In Industry 4.0 environments, machine learning is even considered as the core of PdM, and is advised to be integrated also to business operations in order to link estimations and action recommendations for a better decision making 7.
In the beginning of PdM applications, rule based maintenance might be a good choice to obtain quick business results and to give a stepping stone into machine learning as it is viable both for plants having already IoT, and also for those who have not yet designed their system, but, for an optimum application to realize smart manufacturing, rule based method may not be sufficient if IoT is not implemented, and as historical data will not be available or not continuous in this case.
Among the “Industry 4.0” criteria, two key criteria in context with PdM are “technical assistance” and “decentralized decision-making”. PdM drastically improves technical support by catching errors that humans cannot see . Especially if based on machine learning, PdM decisions are based only on data so that centralized decision making is eliminated. This may not be always acceptable in all types of establishments in earlier steps.
PdM plays a key role in Industry 4.0. With its help, “manufacturers will have full lifespan use of parts and no unplanned downtime will be faced”. This means facilitating smart factories that enable machine to machine learning improving at the end safety and productivity 8.
Additionally, if the technologies enabling PdM are explored, one can see that those technologies also contribute to digital transformation which is necessary for an Industry 4.0 plant. Figure 3 gives a table of technologies that enable PdM 9.
Figure 3: Technologies Enabling PdMSource: Deloitte analysis. Deloitte University Press. https://dupress.deloitte.com
Improving effect of Industry 4.0 on Predictive Maintenance
In real applications, impacts of Industry 4.0 and PdM on each other cannot be separated actually. There is a loop which is always repeating itself after having closed, resulting a continuous development fed by the impact of PdM on Industry 4.0, and vice versa. In other words, PdM fits into Industry 4.0 as Internet of Things (IoT) the main element of Industry 4.0 is a crucial enabler for it 8.
Continuous data from live sensors, and intelligent systems analysing that data are crucial for PdM. This is just the essential point to focus. The more Industry 4.0 tools are implemented, the higher will be the performance of PdM. Industry 4.0 is shortly defined also as “the superposition of several technological developments related to Cyber Physical Systems (CBS), Internet of Things (IoT), Internet of Services (IoS), and Data Mining (DM)” 4. Such a superposition will provide ideal data feed to PdM for estimating, monitoring and controlling of eventual failures and inconsistencies.
Industry 4.0 enables to realize the “smart factory” idea, and smart factories can lead even to “self-aware and self-maintained machine systems which can self-assess its own health and degradation, and use also information from other peers for smart maintenance decisions” 4, which can be considered as an exact fit for industrial big data environment.
For mechanical systems, self-awareness means “being able to assess the current or past condition of a machine, and react to the assessment output 10. The smarter the factory is, the higher will be the self-awareness of the machines, and smartness level is proportional to implementation and application percentage of Industry 4.0 tools.
Mutual interaction of Industry 4.0 and PdM is open to improvements and development as digital transformation has not reached its limits. Tools such as big data environment and cloud computing environment are considered to further facilitate Industry 4.0. Similarly, to facilitate the interaction, studies and applications to further increase the power of PdM are also presented recently.
Such a tool attracting attention is Smart Predictive Maintenance (SPdM) discussed widely in production environments. SPdM is a modern maintenance strategy beyond PdM, and incorporates PdM with several technologies and maintenance systems (e.g. CMMS, ERP, MES). Its main feature is the capability to provide also information on maintenance planning, spare parts planning, and automation of maintenance tasks 11.
We conclude shortly that Industry 4.0 promotes PdM and smart manufacturing as machines are connected as a collaborative community in smart factories generating a great potential for PdM 4. Transition to Industry 4.0 is a dynamic process and continues. PdM plays a fundamental role in that transition, benefitting also from the developments of that transition in terms of data collection and data analysis.
1Alfin, F; Mill Maintenance Methods, Miller Magazine, No.101, May 2018
2Yalçınkaya, A; Etkin Bakım Sistemi, El Kitabı/Kılavuz, AB Holding, Konya, Mayıs 2013
3Orhan, İ; Karakoç Hikmet; Bakım Yönetim Süreçleri ve Etkinliğin Değerlendirilmesi, Mühendis ve Makina, Sayı 607, Ağustos 2010
4Li, Z; Wang, K; He, Y;Industry 4.0-Potentials for Predictive Maintenance, Proceedings of the 6th International Workshop of Advanced Manufacturing and Automation (IWAMA), October 2016
5Chesworth, D; Industry 4.0 Techniques as a Maintenance Strategy (A Review Paper), www.researchgate.net/publication/322369285 Access: 9 June 2020
6Why Predictive Maintenance is Driving Industry 4.0, Smart Industry, White Paper, Seebo Interactive Ltd, 2018, www.seebo.com/iot-resources, Access: 9 June 2020
7Becks, A; Machine Learning als Kern der praedikativen Wartung, IT & Production, Oktober 2017
8Roubaud, J; How Predictive Maintenance Fits into
Industry 4.0, www.engineering.com, Access: 9 June 2020
9Coleman, Ch; Damodaran, S; Chandramouli, M; Deuel, E; Making maintenance smarter, Deloitte University Press, Deloitte Development LLC, 2017.
10Lee, J; Kao, H-A; Yang, S; Service innovation and smart analytics for Industry 4.0 and big data environment, Proceedings of the 6th Conference on Industrial Product-Service Systems, Procedia CIRP 16 (2014).
11Durmuş, M; Smart Predictive Maintenance: Der Schlüssel zu Industrie 4.0 www.aisoma.de , Access: 6 May 2018, and 9 June 2020.