Article first posted at Conscious Reliability by James Reyes-Picknell, Jesus Sifonte, and team.
The Phonograph came first in 1877, cassette players arrived in 1971, CDs made it to the market in 1982 and the just arrived ultra-modern solid state hard drives are devices all capable of recording sound. Likewise, the 1876 telephone and the 2 decades old cell phone are all good for making possible long-distance conversations. Certainly, sometimes technology changes faster than our ability to adapt to it. When you are still getting used to a specific computer operating system X.1 its creator is already announcing the launch of the newer version X.2 and the extra features it brings making it a better option than its predecessor. A change to the new version’s ‘toy’ always seems more convenient than keeping the current stuff – just think of Apple’s product development trajectory!
I heard the definition of the term computer for the first in 1986 while studying my engineering career. My Fortran professor defined it as “en electromagnetic box capable of performing calculations some thousand times faster than humans”. He continued: “With computers you get results (either good or bad) much faster than by doing manual calculations”. The truth is that results are gotten million times faster for very complex algorithms that are almost impossible to be solved by humans manually. We often use computers in the maintenance and reliability engineering field to make calculations regarding our assets’ reliability or even to predict the time for occurrence of certain failure modes. Is prediction more important than planning or good maintenance management practices? Furthermore, what is the benefit of having fast useful life estimations if we don’t have the right tools, knowledge and best practices to extend it to its full potential. Please, fellows don’t get me wrong! I am a firm believer in good technology, especially when it is used to make engineering calculations and predictions with verifiable field data.
Back in the 90’s, many companies moved from Preventive to Predictive Maintenance because PdM was the latest fashion trend in maintenance. Some of them could attain good results by using a reasonable PM and PdM task mix in their programmed maintenance system. Usually, companies complying with their PM program schedules through good planning and scheduling were even more successful after incorporating predictive tasks into it. But, those with poor management practices, failing to comply with their prior PM program, could not succeed with their new PdM engagement. I was called once by a manufacturing company to perform an assessment of their PdM program. The plant’s maintenance manager was not satisfied with the results obtained out of their PdM program since they were experiencing too many unexpected failures. Unsurprisingly, there was nothing wrong with their technical personnel’s knowledge and their ability to detect potential failure with their predictive technologies. They had state of the art tools and adequate knowledge. The organization lacked effective planning and precision maintenance practices needed to backup PdM tools for ensuring the asset’s components long useful life under their current operating context.
Today, IIoT (Industrial Internet of Things) is the latest trend in Physical Asset Management and companies are being attracted to implement it. Certainly, IIoT will help companies to perform better provided they have mature asset and maintenance management systems required for ensuring data coming from automated collection systems is correct, accurate, complete, fit for purpose and easily converted to useful information. At the end, data analysis made from field data is used to make management decisions impacting value realized from assets. Fast processing and accurate analysis of good data will potentially render optimum value from assets to your organization. Likewise, acquiring wrong data at a fast pace will accelerate low performance and result in potentially bad decision making.
Consider reviewing the following “to do list” for ensuring your organization is ready before embarking in a “promising” big scale “magical” IIoT project:
1- Understand your business corporate goal
Economic and Non-Economic corporate goals must drive all improvement initiatives. Understanding how our people, processes and assets contribute towards attaining them is the very first logical step toward obtaining long term and sustainable results from any big scale project. Microphones aren’t more important than voices which might be the principal asset professional singers possess. Similarly, transducers are not more relevant than company goals. But, they may play an important role in providing high quality data for important management decision making for attaining goals. Team up with other departments within your organization to ensure that all agree on short and long-term business objectives. Then, perform an overall gap analysis aiming your people skills, processes and assets to establish priorities. Have a multidisciplinary team ranking your assets through and Asset Criticality Analysis (ACA) per their relevance on attaining company goals.
2- Understand how your assets failures affect lifecycle profits
Estimate failure event risks precisely for your critical assets. Understand what causes them to fail and how each probable failure events affect business goals in any form. Physical asset components may fail simply due to normal aging, which may vary according to the intensity of their operation. Other failures occur randomly. Some of them are evidenced by repetitive operational or maintenance mistakes likely due to lack of training or appropriate tools and working procedures. Understanding failure mechanisms is quite helpful to eventually assign appropriate tasks to reduce their risks to tolerable levels. Not all failure situations are fixed with maintenance tasks. Yet, less maintenance and more training is needed to prevent most of them from happening. Thus, identify your personnel (both maintenance and operations) training, tools and procedure needs to avoid those assets failures caused by the lack of them.
3- Identify Appropriate Failure Consequence Management Policies and Tasks
Some failure causes may require proactive (Preventive or Condition Monitoring) tasks to avoid further failure incidents. Others may only need one-time actions such as physical asset, procedures or work practice modifications to mitigate their impact. When Condition Monitoring (C) and/or Detection (D) tasks are chosen, some quantitative or qualitative parameter or condition is either measured/trended or corroborated. In the case of D tasks, you just corroborate the machine halts when pushing the emergency stop button or when reaching an overload condition. We are just corroborating if the asset has a hidden fail condition. On the other hand, we can apply C tasks upon finding some measurable variable which values change gradually as the failed condition worsens. Condition Monitoring is considered technically feasible if a potential failure condition is found well in advance of the functional failure taking place. Make sure your people understand what failure events they are trying to predict and how they affect business goals before measuring anything, either manually or automatically.
4- Understand Failure Mechanisms First
Measured parameters coming from predictive technologies and process variables are defined once actual cause of failure events are understood. Thus, measurements target specific failure events. P-F times and cost data will help your team calculate measurement intervals to avoid both excess or lack of information needed for failure detection. Proper monitoring parameters selection comes from understanding failure causes – many people think it is the other way around. Automatic data collection and diagnostics can be implemented at its full extent when this process is well understood and put into practice.
5- Implement Precision Maintenance
To implement precision maintenance, you must be able not only to detect bearing wear, but to increase their useful life too. Precision lubrication, through best lubrication practices on lubricant storing, handling and application will ensure longer bearing (and other lubricated components) life. Proper shaft alignment, mechanical clearance adjustments, gear backlashing set ups, rotor dynamic balancing and torque application will increase the components life by a great margin if these practices are not fully implemented yet. It will also contribute to a reduced power consumption impacting the business bottom line as well. Less than optimum results are attained when only detecting alone rather than proactively addressing failure causes.
Understand your company goals. Get your team ready to pursue them. Then, learn how your assets, processes and people skills may affect achieving them –
• Visualize how failure events may occur and in which way they are relevant.
• Filter out critical failure events for which measurable parameters reveal potential failures. Choose appropriate parameters for trending.
• Utilize the right tools and procedures when performing repairs.
• Use precision maintenance and proper maintenance management practices to ensure components useful life is enhanced and used to its full extent.
At that point, consider implementing IIoT thru on-line monitoring and automated diagnostics to enhance your potential failure detection capabilities for improved overall business result.