Predictive Maintenance

The machine knows what condition it is in and can call for help at an early stage. Predictive Maintenance considered an entry ticket to Industry 4.0. Nevertheless, many companies still not using this technology.

Maintenance Strategy of the future?

Europeans manufacturing industry are facing strong headwins thanks to growing operational costs and pressure from outside Europe. Struggle to integrate new innovative technology solution are limiting the potentiall growth of these organizations. Especcialy in such markets, improving operational efficencies and therefore, cutting the operational costs, becomes imperative for most CEO`s in this industy.

Currently existing maintenance processes are fare from efficient and leave plenty of room for improvement. Because of these poor processes, many companies started to use predictive analytics to unlock the streams in data coming from other machinery and turn this data into value. Due to predictive algorithms, companies may be aware at what time their machine will potentially fail.

More than 90% of the companies describe their existing maintenance processes as not very efficient, but are they ready to streamline them? (Milojevic, 2019)

1. Impact of Maintenance

Maintenance costs are a major part of the total operating costs of all manufacturing or production plants. Depending on the specific industriy, maintenance costs can represent between 15 a 60 percent of the costs of goods produced. For example… maintenance costs for iron and steel, pulp and paper, and other heavy industries represent up to 60 percent… (Mobley, 2002)

Surveys of maintenace indicates that one-third, so 33 cents out of every Euro, of all costs wasted due to poorly maintenance.  In the cost structure of maintenance, the consumption of goods or the use of resources for the function of maintenance expressed on a monetary level.

According to the companies, the largest cost block of maintenance was staff costs, averaging 43%. The most pronounced share of personnel costs was almost 48% for the manufacturing companies. The investments in plant maintenance (39%) and infrastructure maintenance (37%) were slightly lower. (Brinkmann, 2000)

If we consider that the European industry spends millions each year for maintenance, there is a high loss every year. According to our research, the result of ineffective maintenance management represents a loss of nearly 50 billion EUR in 2018.

However, why is there an ineffective maintenance management at all?

One of the dominant reason is the lack of actual data to quantify the actual need for repair the machinery. Many companies nowadays are still working with the approach „newer change a winning team“. Every day we rely on a wide range of machines, from the coffee maker to our transportation into our work. Without the right maintenace, failure will occur or eventually these machines will even break down.

2. Maintenance Methode

„Maintenance is a necessary evil“ was always the biggest statement of all. However, in the 21st century, with the use of Artificial Intelligence (A.I.), we have a technology providing the right data to reduce or even eliminate unnecessary repairs and prevent machine failures.

Beside the fact that implemnting predictive maintenace helps companies to reduce downtime and optimize spare parts inventory, the biggest impact is on the maximazation of the equipments lifetime.

To understand the effectifnes of predictive maintenace, Table 1 shows the different strategies and their opportunities.

Table 1 Maintenance strategies and their opportunities

High safety for userX
Higher product qualityX
Optimized maintenanceX
Higher machine lifetimeXX
Reduced energy costsXX
Reduced Human ResourceXX
Low Investment costXX

2.1 Reactive Maintenance

The logic of reactive maintenance or „run-to-failure“ management is straightfoward. Machines are used until they break down, and will be fixed afterwards. This „don`t spend any maintenance cost on a running system“ approach has been there since the first manufacturing plant was build. Few plants are using this „no-maintenance“ approach. Almost every company performs a basic preventive task, like lubrication or adjustments, in their reactive maintenance strategy. However, this strategy brings the highest expenses in:

  • Spare part inventory cost
  • High over time / labor costs in breakdown time
  • High machine downtime
  • Low production availability

Analysis of maintenance cost indicates that a repaire prtformed in the reactive … mode will average about three times higher than the same repair made within a sheduled or preventive mode. (Brinkmann, 2000)

The highest risk with a run-to-failure management approach is the safety issue. Running a machine until it breaks down is a major issue for the operator on the machinery. Figure 1 illustrate the typical reactive maintenance approach.

Reactive Maintenance (MathWorks, 2019)
Figure 1 Reactive Maintenance (MathWorks, 2019)

2.2 Preventive Maintenance

Many organizations try to prevent failure before it occurs by performing regular checks on their equipment. Preventice maintenance are always time-driven programms and therefore tasks are based on elapsed time of operation. Figure 2 illustrates an example oft he statistical life of a machine-train. This mean-time-to-failure (MTTF) indicates the probability of failures of a machine. Failure may occure during the initial installation process and the probability of failure is low during normal life. To reduce the sharply inreased propability with elapsed time, machine repairs are sheduled based on the MTTF statistic.

Typical MTTF (Brinkmann, 2000)
Figure 2 Typical MTTF (Brinkmann, 2000)

The biggest challenge with preventive maintenance is to determin when to do maintenance. If the maintenance is sheduled to early, wasting of machine lifetime, where it is still usable, and costs increase. On the other hand, if the process is sheduled to late, safety-critical factors may arise and the machine eventually could break down. Figure 3 shows a typical preventive maintenance approach.

Preventive Maintenance (MathWorks, 2019)
Figure 3 Preventive Maintenance (MathWorks, 2019)

2.3 Predictive Maintenance

With predictive maintenance, time-to-failure of a machine can be estimate. There are many defenitions for a predictive maintenance approach, to few it`s monitoring of vibration of ratating machinery in attempt to detect incipent problems and prevent catatsrophic failure. On the other side, the monitoring of infrared images of electrical parts such as switchgears or motors is also a part oft he predictive maintenance approach. The common premise is to regular monitor the actual mechanical and electrical condition of a machine and compare it to data such as operating process and efficiency.

However, predictive maintenance is much more.

Predictive maintenance is a philosophy or attitude that, simple stated, uses the actual operating condition of plant equipment and systems to optimize total plant operation. A comprehensive … management program uses the most cost-effective tools … to obtain the actual operation condition of crital plant systems and based on this actual data shedules all maintenance activities on an as-needed basis. (Brinkmann, 2000)

With this type of maintenance a organization optimized the availability of it`s machinery and greatly reduces the cost stucture of its maintenace process. Therefore, this system also improves the quality of products, productivity and profitability of any production and manufacturing plant.

2.4 Proactive or prevention maintenance

This maintenance strategy lays primary emphasis on tracing all failures to their root cause. After analyzing each failure individually, proactive measures are taken to ensure this failure won`t happen again.

It utilizes all of the predictive/preventive maintenance techniques discussed above in conjunction with root cause failure analysis (RCFA). RCFA detects ant pinpoints the problems that causes defects. It ensures that appropriate installation and repair techniques are adopted and implemented. (Scheffer, 2004)

Additional to the advantages mentioned above, a prevention maintenance strategy also highlights the need of redesign and/or modification of certain parts and equipment to avoid recurrence of such problems.

3. Today’s maintenance

Are manufactures aware of these opportunities, and do they have necessary capabilities in place to implement such a strategy? This study sets out to explore how European manufacturers are approaching predictive maintenance strategy from an investment, infrastructure implementation, and strategy perspective. Studies of equipment reliability problems over the past 30 years shows, maintenance is responsible for 17 percent of production interruptions and quality problems (Mobley, 2002).

3.1 Key Findings

  • 89% of the asked manufacturing companies describe their maintenance strategy as not efficient. Unplanned downtime combined with sudden failures of the aging infrastructure are the major challenges.
  • Only 50% of the companies are at least piloting a predictive maintenance strategy, which means there is plenty of room for improvement.
  • Unsurprisingly, data security and privacy concerns are at the top of the list of inhibitors of predictive maintenance developments for 89% of the companies, but there is a significant lack of internal capabilities as well. The major challenges that directly affect the adoption of predictive maintenance and its success relate to an inability to handle growing volumes of available data, to process these, obtain valuable insight and then redesign maintenance processes based on this insight. Inappropriate available technology and infrastructure is another major inhibitor, which is a prerequisite to make the predictive maintenance reality. (Milojevic, 2019)

3.2 Key opportunities and benefits

For most incumbent manufacturers the initial business case to justify the adoption of the Industrial Internet is based on incremental results in increased revenues or savings. As shown in Figure 4, a survey of the World Economic Forum, indicates that companies are turning to digital technology either to drive down cost or increase top-line growth: (Forum, Industrial Internet of Things: Unleashing the Potential of Connected Products and Services, 2015)

  • 79% of respondents indicate that “optimizing asset utilization” is a “very to extremely important” driver for adoption, while
  • 74% say the same about creating alternative revenue streams through new products and services. Accordingly, the most widely cited application of the Industrial Internet is predictive maintenance and remote asset management, which can reduce equipment failures or unexpected downtime based on the operational data now available. Early Industrial Internet adopters such as ThyssenKrupp, Caterpillar and Thames Water are already reaping these types of benefits. Specifically, Thames Water, the largest provider of drinking and waste-water services in the UK, is using sensors, analytics and real-time data to anticipate equipment failures and respond more quickly to critical situations, such as leaks or adverse weather events. (Devraj, 2019)
Business benefits for driving near-term adoption (Forum, 2014)
Figure 4 Business benefits for driving near-term adoption (Forum, 2014)

3.3 Challenges and risks

Despite the great opportunities of predictive maintenance, there are also many factors hinder future adoption and growth. Figure 5 of the World Economic Forum Industrial Industrial Survey shows results on perceived adoption barriers. Almost two-third agree with the widely held view that security and interoperability are the two biggest hurdles. Not surprisingly, 24% state technology immaturity as an significant barrier.

Within our research, we received more concerns and challenges including:

  • The heavy upfront capital investment
  • The lack of understanding values of predictive maintenance among management
  • The lack of vision and leadership
Key barriers in adopting the Industrial Internet (Forum, 2014)
Key barriers in adopting the Industrial Internet (Forum, 2014)

4. Recommended actions for stakeholders

The continuous developments in big data, machine-to-machine communication, and cloud technology have created new possibilities for the investigation of information derived from industrial assets. Condition monitoring in real-time is viable thanks to inputs from sensors, actuators, and other control parameters. What stakeholders need is a bankable analytics and engineering service partner who can help them leverage data science not only to predict embryonic asset failures, but also to eliminate them and take action in a timely manner. (Otto, 2018)

As our research shows, predictive maintenance is already here, delivering real benefits including improved operational efficiency and measurable business outcomes. However, at the same time we found a number of challenges that could slow down the adoption process in organizations. To seize the opportunities, stakeholder need to take immediate actions to overcome the biggest risks including security and price.

Invest in long-term, strategic R&D is highly recommended for all stakeholders. The future development of predictive maintenance will require large-scale multistakeholder efforts in boosting security, reliability and interoperability, and in delivering large-scale benefits. In particular, security is one issue that no one industry, business or government can solve on its own. To address these challenges, frontline contributors will come from academia and industry. Government agencies will play critical roles as well. (Forum, Industrial Internet Survey , 2014)

After evaluating the different maintenance strategies and having decided to implement predictive maintenance, businesses begin a journey with new learnings and insights waiting along the way. Initially, there is no certainty as to which level of failure prediction can be reached. In our experience, an incremental approach as illustrated in Figure 6 has proven to be valuable for mitigating project risk. Each step yields better results leading to reduced downtimes and increased productivity. (Deloitte, 2017)

Journey towards higher levels of predictive maintenance (Deloitte, 2017)
Figure 6 Journey towards higher levels of predictive maintenance (Deloitte, 2017)

5. Can predictive maintenance help your company?

Predictive maintenance can help you manage maintenance more efficiently. However, keep in mind that not all enterprises require the same level of reliability from their assets. A good place to start the assessment for your enterprise is to look at mission critical requirements and maintenance program maturity. (Deloitte, 2017)

Ask yourself the following questions:

  • How reliable do our assets need to be?
  • What are our availability targets?
  • What is our machines’ current failure-rate?
  • How high are our current maintenance costs?
  • Do we have the right spare parts in the right place at the right time?
  • How do we determine when it is time to replace an asset rather than to maintain it?
  • What data do we already have that is not being used effectively?
  • Have we identified the critical assets in our production system?
  • Are there some critical assets that would benefit from a predictive maintenance program?
  • Do we have the needed technological expertise in house to develop a predictive maintenance program?
  • Do I have advanced analytics experts in house?

Appendix A: About the research

This report is a part of the Schalk Vermögens- und Verwaltungsgesellschaft mbH research initiative „AUX Research Institute“ launched 2019 by our founder to gain free access to knowledge an research material. The research was undertook to establish a clearer understanding of the transformative opportunities a new risks arising from predictive maintenance. Further objectives of the project included helping business envision the new world resulting and identifying potential implications to the workforce (e.g. employment, skill distribution and productivity).

About AUX Research Institute

Founded in 2018, Schalk Vermögens- und Verwaltungsgesellschaft mbH is a independent consulting and research firm for digital transformation and start-ups. In 2019 we launched our new research institute, AUX Research Institute.

With our new Institution, the goal is to publish research on large scale and give access to everyone. Spreading knowledge and helping to educate people around the world.

!Download the articel here!


Brinkmann, P. D.-P. (2000). Kostenrechnung für die Instandhaltung- Ergebnisse einer empirischen Untersuchung. Bamberg: Universität Bamberg.

Deloitte. (July 2017). Predictive MaintenanceTaking pro-active measures based on advanced data analytics to predict and avoid machine failure. Von Predictive MaintenanceTaking pro-active measures based on advanced data analytics to predict and avoid machine failure:

Devraj, A. V. (01. 08 2019). News Room Accenture. Von News Room Accenture:

Forum, W. E. (2014). Industrial Internet Survey . Geneva: World Economic Forum .

Forum, W. E. (2015). Industrial Internet of Things: Unleashing the Potential of Connected Products and Services. Geneva: World Economic Forum.

MathWorks, M. . (05. 07 2019). MATLAB – MathWorks. Von

Milojevic, F. N. (09. 07 2019). sitsi market research service. Von sitsi:

Mobley, R. K. (2002). An introduction to predictive maintenance. Amsterdam: Butterworth-Heinemann.

Otto, S. (12. December 2018). How Predictive Maintenance is Improving Asset Efficiency. Von How Predictive Maintenance is Improving Asset Efficiency:

Scheffer, P. G. (2004). Practical Machinery Vibration Analysis and Predictive Maintenance. Burlington: Newnes.

Recommended Posts

No comment yet, add your voice below!

Add a Comment

Your email address will not be published. Required fields are marked *