Select Page

The following is an abstract from a new research project on the adoption of Maintenance 4.0 and Machine Learning based Predictive Maintenance. The study was conducted by Emory University students and sponsored by Presenso.

Research Objectives

The objective of these questions was to gain insights into the following:

1) The extent to which O&M employees believe that their organizations have the infrastructure and expertise to deploy Maintenance 4.0.

2) There is buy-in from senior management for Maintenance 4.0

3) The inhibitors or blockers of Maintenance 4.0 deployment


A lack of understanding of Industry 4.0 was cited as the strongest blocker of deployment (4.0/5.0).

The second most significant factor preventing deployment was a skills shortage of engineers (3.8/5.0).  Furthermore, when asked to rank statements related to deployment readiness, the lowest ranked statement (4.6/9.0) was that “we have sufficient staff of data scientists to deploy Maintenance 4.0.”

Maintenance 4 Blockers

From a readiness perspective, most O&M employees believe that their organization has the core infrastructure in place needed for Maintenance 4.0. (7.0/9.0 ranking).


O&M employees’ concerns about the lack of data scientists within their organization was raised in both the online survey and interviews. It should be noted that the vast majority of O&M employees have not been exposed to topics such as Machine Learning and Artificial Intelligence.  These subjects were unlikely included in their academic or work-sponsored continuing education.

To some extent, this O&M feedback reflects a fear of the unknown. In reality, advances in Machine Learning and other technologies are enabling deep analytics to be performed offsite without the input of plant level employees. Therefore, it is likely that these employees are unfamiliar with advances in analytics solutions and are concerned about the relevancy of their skillets.

In limited cases, when Digital Twins of existing machinery is developed, there is a need for the expertise of those plant technicians familiar with the operations of the underlying asset.  However, the Digital Twin is unlikely to become the default Predictive Maintenance solution based on the relatively high deployment cost.

The relatively high agreement with the statement “my organization has the core infrastructure needed for Maintenance 4.0” was surprising.  It reflects that the view that technical considerations are less of a constraint than people/process issues.

Finally, it is interesting to note that the statement “facility staff recognize the potential of Maintenance 4.0” scored higher (5.6/9.0) than the statement “senior executives recognize the potential of Maintenance 4.0.”  Other third-party research indicates that senior management is both aware and committed to Maintenance 4.0.  Apparently, this perception has not been fully communicated at a plant-floor level.


The concerns about a lack of data scientists should not be discounted.  These issues need to be addressed directly, even if they are not completely warranted.  As we have discussed previously, it is important to appoint internal change champions and evangelists who can create buy-in at a plant level.  This role will require communicating both the benefits and challenges of Maintenance 4.0.  There are many different routes to adopting Machine Learning based Predictive Maintenance – the assumption that existing employees will need to re-skill is not necessarily accurate.

Further Research Information

Maintenance 4.0 Research Infographic

2018 Emory Research Project:  The Future of IIoT Predictive Maintenance

Further Reading and Guidance from Presenso

Are O&M Employees’ Concerns About Maintenance 4.0 Justified?

How to scale Machine Learning-based Predictive Maintenance solutions

Is the Smart Factory a Good Investment? An Incremental Approach to Digitalization