ramwright_3eom1p

We’ll further the discussion from our recent post “DON’T BASE YOUR RELIABILITY STRATEGIES ON YOUR ORIGINAL ESTIMATES ONLY. YOU HAVE TO BE “DYNAMIC”!”

Last time we talked about moving Pf on the Pf-Ff Curve and why that happens.

Today, I’d like to explain how the entire Pf-Ff Curve changes dynamically to move left and right on the X-axis (Age). In the figure below, we have both the Pe-Fe Estimate Curve (Grey) and the Pf-Ff Forecast Curve (dotted). The slope at the current point on the estimate curve is compared to the current slope of the forecast curve over a short interval toward the past; this factor is user selected. Another slope is measured from Installation to Now. Through use of an algorithm that mathematically compares these slopes against each other to determine a value called the “Decay Rate”. Depending on whether the Decay Rate is positive or negative, the forecast curve may move toward the RED High Stress curve, or toward the GREEN Low Stress curve respectively. As time or cycles pass by, the Pf-Ff Forecast Curve becomes more and more confident and thus, Point Pf (Potential Failure Forecast) becomes more stable and permanent in its location. Barring any catastrophic external anomalies resulting in failure, we will now be able to plan and schedule our maintenance activities.

#dynamicpfcurve, #pfcurve, #assetmanagement, #conditionmonitoring, #predictivemaintenance

I recently was collaborating with some folks over on the Sustainable Asset & Maintenance Management (A&MM) LinkedIn page regarding the DPFC and how it is better than just the traditional P-F Curve. I’ve listed the relevant post below.

Part 3: The Dynamic P-F Curve. The example of the milk carton emptying can be modeled with our DPFC and the rate of change of the Estimate vs the Forecast P-F Curve. The curve shown in marker on the carton is the Estimated P-F Curve. The reality is that when the boys have a sleep over and there are now 3 boys instead of one drinking from the carton, the original P-F Curve is useless to prevent Functional Failure. The estimates are based upon past history, not current conditions. The volume that is now leaving the carton has changed and therefore, the P-F must also change. IT MUST BE DYNAMIC. This is the basis for our Dynamic P-F Curve.

Sustainable Asset & Maintenance Management (A&MM) on LinkedIn

This post was added as from a LinkedIn question to me regarding expiration rather than level. I thought it was a good enough though experiment to add to this post.

LinkedIn Post Feed Link

This year Gary will be presenting at RELIABLEPLANT 2023 in Orlando, Florida USA. The topic will be Dynamic Procedures and how todays mobile technology and software can help improve the overall Work Management process. Say good-bye to paper!

The Dynamic P-F Curve is not unlike the traditional curve; however, the addition of current information or online data inputs allow the curve to become “live” or “dynamic” through the use of AI and its embedded algorithms, thereby making adjustments based on future Dynamic P-F Curve forecasts.

Today, I’d like to focus on the two points on the curve, the PE and FE points on the solid grey line. These are our traditional points, well known to the reliability community. These are our estimates of Potential Failure and Functional Failure, on the curve, as they relate to an item’s resistance to failure (Y axis) against age (X axis). Their locations are based on well-established industry best practices as well as our own empirical data. PE is selected as the point on the curve at which “failure is imminent”. Merriam-Webster defines IMMINENT as “ready to take place” or “happening soon”. In both definitions, “imminent” is a judgement call of the reliability engineer based on the historical data and an understanding of current operating conditions that the item is exposed to; because of this, PE is entirely malleable in its position on the P-F curve and up to the best judgment of the RE.

Now armed with this understanding, let’s ask the question “Why does the P-F curve even exist?”. The primary purpose of the curve is to establish the criteria by which an item will be taken out of service and scheduled for maintenance prior to functional failure. To ensure adequate time for Planning and Scheduling, the P-F Interval must be of sufficient time. The Dynamic P-F Curve helps to accomplish this by measuring the slope of the Current Asset Curve vs. the Historical Baseline Curve; based on this slope, the algorithm calculates a new PF in real-time and displays the Algorithm Forecasted Curve. By moving PF up the curve, we give Maintenance the time required for an adequate response prior to Functional Failure of the item.

Our next post will cover the “Probable Range of PF- FF”. Thanks for reading; your comments are welcomed as we explore this new and exciting use of the Dynamic P-F Curve.

After 38 months of working with the best group of Maintenance & Reliability professionals I’ve had the pleasure to meet; RAMwright’s project as a sub-contractor deploying Asset Management best practices has ended. We worked within a larger project that included Human Performance Management (HPM), Materials Management (MM), and Asset Management (AM) for Implementation Engineers (IE) APE system. I want to thank IE for their trust in RAMwright to deliver a quality portion of their very important project. RAMwright successfully implemented the SAP PM module, as well as a Asset Criticality Analysis, Maintenance Work Management System, RCA program, RCM program, and a PME/PMO program (including Precision Maintenance Procedure Writing) across three baking facilities. The image below is the leadership team that successfully rolled out this Reliability Implementation; in order from left to right are: Suru Patel, Masoud Niknam, Mehdi Vahedi, Gary Josebeck, Blaine Karr, and Abdul Rafay Khan. Not all of the wonderful folks that participated in this project are represented in this picture, but I want them to know they will be missed.

We also conducted training on other Reliability Engineering topics such as Weibull Analysis, Reliability Block Diagrams, Reliability Growth, Lubrication Excellence, Predictive Maintenance, and Non-destructive Testing. The first image below is the SuperSmith software that we used to evaluate the 3-parameter Weibull to determine Beta and Eta. We also used Weibull paper to determine these parameters manually. The second image below is our Optimum Parts Replacement Tool that is used to determine replacement frequency for Time-directed Tasks. This tool was used to justify frequencies previously selected for PMs; in many cases, we were able to extend the intervals between maintenance activities thereby saving labor resources and money, without adding risk to the business. The third image is our same tool, but used to determine the correct PM frequency based on Costs. Notice that PM Frequency at 300 days yields the lowest Total Maintenance Costs (including PM and Failure Costs).

While it’s always hard to leave new friends and the relationships that we build, I leave feeling confident that the team will be able to carry on the processes that we’ve implemented and coached them on all these years. I look forward to checking in on them in the years ahead! Gary.