Speaking Of Reliability: Friends Discussing Reliability Engineering Topics | Warranty | Plant Maintenance

Demonstration Testing Abstract Chris and Fred discuss the whole idea of ‘demonstration testing’ … and how it may be ‘chronically’ overused or at least misunderstood. Heard of ‘demonstration testing’ and need to learn more? Key Points Join Chris and Fred as they discuss Topics include: What is ‘demonstration’ testing? Typically … ‘demonstration testing’ involves testing a system in ‘real world’ scenarios (no acceleration), without any assumed knowledge, using no models based on previous information. This might sound very ‘objective’ and ‘scientific’ … but it is very expensive and time-consuming. Think about something that is ‘really’ reliable. Say something has an MTBF of 10 years. How long does testing need to be to get enough failure points to be confident about the real MTBF? Who likes ‘demonstration’ testing? Typically governmental, military and some commercial organizations. But do we really ‘demonstrate’ reliability? Not really. For example, we typically do reliability demonstration testing to work out if we are going to purchase a large number of these systems.  Which means we conduct reliability demonstration testing on a pre-production prototype. And what can really affect reliability? Manufacturing. So what is the point of testing a prototype that is not built using the ‘actual’ manufacturing process? Look at the United States Department of Defense (DoD) acquisition of the armored Joint Light Tactical Vehicle (JLTV). It passed reliability demonstration testing really well … it actually ‘demonstrated’ a reliability that was three times higher than the requirement. But now it is being introduced into service … it has a lot of reliability problems. Another example? Ever heard of the Hewlett-Packard (HP) digital camera? No. The prototype pass a reliability demonstration test – but the contract allowed the manufacturer to substitute any component it wanted to at will. And it rarely worked longer than a month. So ‘demonstrating reliability’ is not really the same as making something reliable. Demonstrating reliability is often an arduous activity where we need to pass a test. And this means we aren’t testing to learn. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 685 Demonstration Testing appeared first on Accendo Reliability.

Perils of Assuming a Model Abstract Chris and Fred discuss some of the issues you face when selecting a (potentially incorrect) model. And how much uncertainty still remains when you do. Hate statistics … but need a nice and easy introduction to this sort of stuff? Key Points Join Chris and Fred as they discuss some of the issues and problems we are faced with when we select a model. Topics include: Confidence is a measure of you. Look at the video below! This is what happens in our confidence of a (Weibull) probability distribution when we get more data. The contour lines around the ‘bell’ curve get much ‘tighter’ when we have more data. This shows the statistical effect more data can have on your confidence that you understand the nature of the underlying probability distribution. But you can be ‘fakely’ confident as well. What happens if we assume the wrong model? Let’s do exactly the same thing we did above, but instead of using a curve that aligns with the density of the data points, use a clearly incorrect model. Which in this case is the Exponential distribution. And this podcast is all about confidence … and how it affects everything from what you measure, test, specify and design into your system. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 684 Perils of Assuming a Model appeared first on Accendo Reliability.

Public Reporting of Failures Abstract Kirk and Fred discussing the challenges of failure analysis and reporting of the root cause of electronic system failures and corrective action to the consuming public Key Points Join Kirk and Fred as they discuss product failure reporting to the public. Topics include: Backblaze, a data storage company, publishes the annualized failure rates between thousands of  hard drives from a variety of manufacturers and yet the drive manufacturers do not provide the public the most important data which is the root cause of those failures. Consumer Reports has a spring and fall surveys from the membership on consumer satisfaction including the reliability of products they purchased recently and that is one source of comparative reliability data available. The Warranty Week website does report warranty costs for public companies and that can provide motivation for companies with similar products to compete on reliability and reduce warranty failure rates. Legal liabilities prevent the  public disclosure  of the cause a vast majority of product failures and that limits the broader knowledge of reliability engineering, but this conflict is not going to end anytime soon. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes Here is a link to the website Warranty Week that we discuss in this podcast Please click on this link to access a relatively new analysis of traditional reliability prediction methods article from the US ARMY and CALCE titled  “Reliability Prediction – A Continued Reliance on a Misleading Approach” For more information on the newest discovery testing methodology here is a link to the book “Next Generation HALT and HASS: Robust design of Electronics and Systems” written by Kirk Gray and John Paschkewitz. The post SOR 683 Public Reporting of Failures appeared first on Accendo Reliability.

Heroes versus Prevention Abstract Kirk and Fred discussing why the service departments will get more accolades for repairing failed systems than those in design department that find and prevent failures during development. Key Points Join Kirk and Fred as they discuss the challenges of making prevention of failures a higher value and more visible to management than those field service engineers that get expensive capital equipment running again after failing and stopping manufacturing. Topics include: The inventor(s) of the modern day smoke detectors have saved many more lives than the fireman whose name and picture may be on the front pages of the newspaper and yet few know who invented the smoke detector. A development product team is rarely recognized for finding and preventing failures before market introduction, where the field service technicians that repair the systems are recognized and rewarded. Finding product weaknesses late in the development cycle that delays product launch is a tough call for the one who discovers it and a big challenge to tell management. Sometimes there is a conflict of interest between what a manager is paid for and reliability. If a manufacturing manager is paid for reducing manufacturing cycle time they will not want to increase that time by final stress testing or screening if warranty reduction is not a performance review goal for them. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes Please click on this link to access a relatively new analysis of traditional reliability prediction methods article from the US ARMY and CALCE titled  “Reliability Prediction – A Continued Reliance on a Misleading Approach” For more information on the newest discovery testing methodology here is a link to the book “Next Generation HALT and HASS: Robust design of Electronics and Systems” written by Kirk Gray and John Paschkewitz. The post SOR 682 Heroes versus Prevention appeared first on Accendo Reliability.

Selecting the Right Recommendations Abstract Carl and Fred discussing methods for selecting reliability recommendations, following the process used to develop effective reliability plans. Key Points Join Carl and Fred as they discuss a student question. How do you select the right recommendations? Topics include: Start with reliability gap analysis Consider organizational maturity Consider program decisions and timing Identify reliability methods that need to be implemented Recommend tasks that support the methods Prioritize the tasks, so you are not recommending too much What are the most important criteria for prioritizing tasks? New technology, new applications, new user conditions Safety is always priority Consider cost and time Time to market risk High severity, even with low failure rate, can be high risk Management input is critical Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 681 Selecting the Right Recommendations appeared first on Accendo Reliability.

Pareto was Right Abstract Carl and Fred discussing the subject of prioritizing tasks, both in life, and in the field of reliability engineering. Key Points Join Carl and Fred as they talk about Pareto, his life and contributions. Discussion topics include the 80-20 rule, and how it can be applied in reliability engineering. Topics include: Organizing tasks based on two criteria: time urgency and importance [Note: see Show Notes for reference to a diagram of the “Time Management Matrix”, which describes this concept better than the description in the podcast] Finding time for important tasks that are not time-urgent Pareto: observations about the “80 – 20 rule” Why does the 80 – 20 rule work? Value of prioritizing failures Why should reliability plans should focus on the vital few methods? Focusing extra supplier efforts on critical parts Anticipating the unexpected Parallels tracks: due care on most areas, and extra care on vital few (prioritized) areas Don’t forget about the 20% of issues that are not covered in the 80 – 20 rule How to narrow down which reliability methods are most important for a reliability program Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes The post SOR 680 Pareto was Right appeared first on Accendo Reliability.

Should We Ban the Term Failure? Abstract Chris and Fred discuss the term ‘failure’ … and should we use it? Key Points Join Chris and Fred as they discuss how important words can be … and how they can have a positive effect on behaviors. And sometimes they can just be simple labels that change nothing. Shakespeare said ‘… a rose by any other name would smell as sweet …’ What about ‘putting lipstick on a pig?’ What about the word ‘failure?’ Topics include: Some organizations ban ‘negative’ words because they want to avoid thinking about ‘negative’ things. But failure mechanisms don’t read textbooks, guidebooks, process documents or anything else that describes how we hope they describe. In practice, ‘failures’ are bad. But … they are inevitable. So we can ‘celebrate’ failures that we uncover in our laboratory or CAD package. And management gurus like doing similar things. Simon Sinek advocates that we should ban the word ‘falling’ and not ‘failing.’ He was advocating that we need to make mistakes, embrace risk, and EXPECT issues. He was saying that this is a necessary step en route to success (… or progress). But because ‘failure’ has such a negative connotation … this is (apparently) discouraged in contemporary organizations. Is this an issue with the word ‘failure’ or the ‘leaders’ who own the culture? Reliability engineering needs to talk about failure all the time. And plenty of organizations who do wonders when it comes to reliability use the term ‘failure’ a lot. With clearly no adverse effects. The key is how we treat each one. One failure can be dealt with early in production or eliminated entirely. It is still a failure … just one we never allowed the customer or user to discover. Then there are failures that customers or users find for us. Both are failures. Only one of them is bad. … and there are instances where changing words can have a positive effect. As long as that change in word serves as a constant reminder that your leadership team has changed something. And is doing something different. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 679 Should We Ban the Term Failure appeared first on Accendo Reliability.

Wrapped Around the Axle Abstract Chris and Fred discuss what it means to be ‘wrapped around the axle.’ It is one of those perfect analogies … especially for people who fall for imperfect analogies! Like that assuming that all we have to do is wait for ‘early failures’ to disappear before we get into to the ‘bottom part of the bathtub curve’ … really? Key Points Join Chris and Fred as they discuss how people can get wrapped around the axles in regard to assumptions, axioms, trains of thought and lots of other things that tend to be written in textbooks (or blogs). And if you get too focused on them … you stop focusing on your system. Or reality. Topics include: Early failures sometimes aren’t early failures. Just because you have failures occurring earlier than you would like, doesn’t mean you can assume that they are ‘infant mortality or ‘wear-in’ failures. You know … those failures that tend to ‘work themselves out’ of a system. Minimal usage doesn’t mean minimal damage (or wear). Sometimes vehicles (like taxis) that are driven a lot last for millions of km or miles. But ‘old’ trucks that have low milage are often a mess. Why? Because machines are designed to be used. And if they aren’t used, this means unanticipated failure mechanisms become dominant. The bathtub curve is rarely the bathtub curve. A lot of systems have wear-in, constant hazard rates, and wear-out regions. But that doesn’t mean they have a hazard (or failure) rate that looks like any bathtub you could recline. Some so-called bathtub curves look more like ‘ticks’ or ‘checkmarks.’ Outliers are rarely outliers. Outliers are data points that we exclude from consideration because of bias. And we mean ‘bias’ in terms of what you think or believe should happen. If you remove a data point just because it doesn’t ‘fit’ or is too far ‘away,’ then you are quite literally introducing a bias toward the points you haven’t excluded. Which doesn’t mean anything is an outlier. … so what do you do? Think. Research. Google Scholar. Libraries. Asking questions. Explaining what this means in a simple message that matters to the decision-maker. Explain that assumptions could cost $ X million. You never know! Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 678 Wrapped Around the Axle appeared first on Accendo Reliability.

Military Customers, Center of Gravity, and Reliability Abstract Chris and Kirk discuss how militaries go about making and getting reliable stuff. And this is sometimes more successful than others … Key Points Join Chris and Kirk as they discuss how military equipment is ‘made’ reliable, or procured in a way that promotes reliability. And regardless of which country you look at, there are a number of similarities that seem to become remain consistent. Topics include: Let’s start with MIL-HDBKs and MIL-STDS. All of which have been removed from circulation (technically). But are still routinely referenced in contracts and reliability programs. For example, MIL-STD 217 has a bunch of failure rate estimates for all sorts of components – especially electronic components. But (apart from the fact that it isn’t possible to cast a single failure rate across an entire market) … this document hasn’t been updated for 25 years. How far as your cell now smart phone come over the last 25 years? Are these electronic components anything like what they were in the 1990s? Conflict of interest? Ahh … yes. The majority of profit of contractors who provide equipment make most of their profit from servicing and maintenance. Which means that if they make something more reliable, they make less profit. Reliability engineers can tell you many (informal) stories of how contractors outright admit that they aren’t interested in making their system too reliable as this diminishes profit later on. Customers and culture? Let’s say you have a garden, and you hire a gardener to look after it. Let’s also say that gardener doesn’t do a good job. Would you pay that gardener to come back to do some more gardening … until it is right? Of course not! That is reinforcing a behaviour. But that is what happens in military contracts across the world. A contemporary example is the F-35 program. The reality is that there is a whole ‘military or defence industry.’ The common denominator is that the customer is ‘military or defence.’ So … what is the reason we have this industry? Then there is HALT, HASS and FMECAs. All of them are contained in MIL-HDBKs and MIL-STDs. But are simply aren’t done. FMECAs for example have devolved into massive spreadsheets of failure rate estimates that are identified after the design has been completed … where FMECAs are supposed to be done early to inform design! Contractors respond to what the customers ‘say’ they value. And this comes down what is said at design reviews, contract meetings, briefs et cetera. If the customer is all about milestones and budget … then that is what the contractor is forced to focus on. What hurdles to military customers face? First … they have stopped being ‘engineers.’ It has been a long time since they designed any equipment ‘internally.’ This makes it challenging to oversee contractors do it right. Then there is turnover, where military project managers (senior officers) are routinely posted out of that position. Military customers also struggle to prioritize failures. ‘Every failure is critical.’ Therefore no failure is critical. Nor is it understood how much each failure costs. Militaries also tend to not focus on engineering stuff … at all. Project managers are often cycled in after they have graduate from staff colleges and other institutions that focus on tactics and battlespace stuff. And finally, there are the ‘buckets’ of gold. There is no motivation to spend money from ‘bucket of gold A’ to save three times that mount in ‘bucket of gold B.’ Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 677 Military Customers, Center of Gravity, and Reliability appeared first on Accendo Reliability.

FMEAs and Percentage Reduction in Failures Abstract Chris and Carl ask each other … by how much do FMEAs improve reliability? … or reduce failure? Key Points Join Chris and Carl as they discuss what benefits you can expect from doing a Failure Mode and Effects Analysis (FMEA). And by benefits … what fraction of percentage of failures can you expect a FMEA to prevent? Now if you have never heard of a FMEA – find out about them! They are wonderful group activities where we stop and think about how things can fail … and then come up with corrective actions to prevent failure. Topics include: We sometimes need to convince bosses. And they often want to know what a FMEA can do for them. The first thing we should focus on is that FMEAs prevent design problems, production issues, and all sorts of other things that blow out budget and schedule. And of course, FMEAs can help us focus on the more ‘severe’ failures first. Which might make bosses more interested. … but there is no ‘guaranteed level’ of failure reduction when you do a FMEA. There are lots of different things that effect how effective FMEAs are. And perhaps the most important factor is how seriously the organization takes FMEAs. Not just how serious one or two ‘FMEA warriors’ take it. Organizations that take FMEAs seriously have a training regime, agreed procedure, refresher training, and all sorts of other things that make FMEAs amazing. And of course, if you have been improving reliability year after year, model after model, you might not have a lot of failures to prevent! some FMEAs reduce failures to a ‘level of acceptable risk.’ But this is subjective. And there are lots of really good ideas and corrective actions that should be incorporated even if risk is already at a ‘level of acceptable risk.’ and don’t forget that FMEAs can also be used to identify key design characteristics that are all about customer experience. So it isn’t always just about failure reduction. SO WHAT IS THE NUMBER? We know it depends on lots of different factors, but it is not unreasonable for a FMEA on the first prototype of a brand new product to reduce (in our opinion) a 90 % reduction in failures and production issues. But if you have a ‘mature’ product, then a FMEA might reduce potential failures by perhaps 10 %. But for a ‘mature’ product that is a market-leading device, a 10 % increase in reliability could be all the difference it needs to be considered cutting edge. Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches. Download Audio  RSS Show Notes   The post SOR 676 FMEAs and Percentage Reduction in Failures appeared first on Accendo Reliability.