Reliability Engineering in the Era of Predictive Analytics

Predicting the Future 

Industrial operations are changing in real time. Before we can discuss how to improve the physical management of assets through digital transformation, we must first understand where and how an asset reliability management program creates value. Only then can we understand how investing in digital tools can enhance and accelerate improvements in the future. In our view, all of the value of asset management originates from two sources, the application of Risk Based Maintenance (RBM) methods and the implementation of better Work Management practices.  

Given that value from an asset reliability management program comes from just these two sources, we then need to understand how IIoT and Data Analytics help capture that value above and beyond traditional techniques and tools. In any case, IIoT and Data Analytics are not a magic bullet that will somehow unlock millions of dollars of hidden value. Rather, these are merely tools and, as with any tool, do not create value without having good process and practices underlying their usage.

Proactive Maintenance Optimization

Reliability engineers provide value by applying a risk based approach to reliable maintenance. Value-added activities include developing an accurate Master Asset List, performing spare parts and root cause analyses and ranking asset criticality to determine the priority and maintenance requirements for each asset. When viewed through the lens of RBM, organizations are more aware of which assets are most critical—requiring preventive and predictive maintenance—and which are noncritical, or run-to-fail.

RBM strengthens enterprise asset management in the following ways:

• Eliminating maintenance on non-critical assets, thus saving labor and parts
• Tuning PM intervals on critical assets, thus optimizing cost
• Identifying impending failures on mission-critical assets, which gives enough time to plan, kit and schedule corrective maintenance with minimal business impact

Work Management

Work Management means the process and practices by which work is performed. These activities include work identification, planning, scheduling and execution, as well as outage and contractor management. Work management reduces the number of labor hours needed to do the same work because the client has control of the work through:

• Centralized, consistent planning and kitting of jobs as opposed to the ad hoc approach still favored by many reactive organizations
• Scheduling of work based on production schedule as opposed to ad hoc and uncoordinated approach still favored by many reactive organizations

What is the Value of Improved Predictive Ability?

Looking closer at the value that IIoT and Data Analytics provides, their role can be limited in the reliability and maintenance of noncritical or critical assets. Since noncritical assets are run-to-fail, there is no value from improved predictive ability. When considering critical assets, the PM intervals can be tuned with a trigger set in the data historian and an electronic interface with the EAM system. For example, the differential pressure across a filter saves a visual filter check but needs a trigger to be sent to the CMMS to generate a filter replacement work order.

However, there is a potential in applying IIoT and Data Analytics for managing mission-critical assets. Here the question is whether Conventional Predictive Maintenance (PdM) techniques are providing sufficient advance notice of impending failures to give enough time to plan, kit and schedule the corrective maintenance with minimal business impact. If not, then more advanced IIoT tools may have a role to provide earlier notice based on the ability to detect critical process parameters' anomalies.

Another consideration is whether there are IIoT applications to detect failure modes earlier than conventional PdM techniques which could be applied to detect deviations in process parameters that would correlate to asset component health. For example, a deviation in particle size may be an early warning sign of a deteriorating mill head. Another example is a machine-learning model mapping real-time hydrocarbon readings against a standardized set of operating ranges for a gas compressor. Once the smart sensors detect a deviation from the predicted operating ranges, the system immediately alerts the user and provides a likelihood of compressor failure. In these cases, digital tools help to optimize assets and support the evolution of a reliability culture and strategy.

Our Approach

While digital tools certainly enhance traditional services, we believe that without the essential asset reliability management business process fundamentals in place, these solutions may fail to deliver on all of their intended benefits. In order to maximize return on investment and realize the full potential of digital service offerings in use today, a Reliability and Maintenance Engineer provides the process knowledge, guidance and experience needed for a more strategic application of digital technology.  

"The role of the Reliability and Maintenance Engineer is even more critical in the Digital Era because knowledge and experience are what provide the critical link between data analysis and insight-driven action."

GenesisSolutions' approach is based on the proven engineering principles of asset reliability and maintenance, which we apply to help organizations optimize the latest digital tools and improve on work management practices for more effective asset management. In the end, achieving optimal performance from more reliable and efficient assets requires a proactive maintenance strategy from traditional expertise as well as the predictive capabilities of the latest generation of digital solutions.