Mastering 8D Problem Solving: Root Cause Analysis (D5-D8)

1. About the Course: Hello, everyone. This is the introduction lecture of course and I will make a quick summary for you to know what awaits you. At first, we will start with the root cause analysis, including fishbone, FDA and 55, which is the most important part of this course because all the next steps will be connected with that root cause analysis. Therefore, I just want you to make extra focus for this root cause section. Just in the next step D six, we will start to develop corrective and preventive action plans and we will understand their differences, how we are addressing these actions according to the root cause. In the same step, we will also start to review and update PFMEA, process failure mode and effects analysis, which is another key point of this course. In the seven, we will verify and validate these corrective and preventive actions with some techniques like reproduction of defects. At the last step D eight, we will start generating lessons learned cards and following at colosre Audit and recognizing and celebrating at the end of each step, we will have case study to reinforce our learning with practice. We will go through with one problem to understand the logic between all the steps. For example, in D five, we will make a root cause analysis for a problem in the following step, D six, we will have actions against the root cause that we have done in the previous step. If you are ready, let's start. 2. Overview the 8D Steps: 80, it stands for the eight disciplines, a structure problem solving methodology. In D one, we identify the problem using the 5w2h methodology. In D two, we analyze similar parts and processes to ensure they aren't at risk. In D three, we conduct an initial analysis, construct non detection points, and check the basic process conditions to ensure everything is in order. In D four, we develop a containment action plan based on the data from the previous steps. This is an important step because it is very extinguished the fire and reduce the tension in the working environment. Now in this course, we will begin with the root cause analysis. We will use effective tools here, including the fish bone analysis, faulry analysis, acronym is FDA and five fs root cause analysis. This step is a milestone in problem solving because all the subsequent steps depend on the outcome of this analysis. In these six, we will focus on corrective and preventive actions. This is where we develop counter measures based on the identifier root causes. We will understand the difference between the corrective actions and the preventive actions. In this step, we will also learn how to review and update PFMEA, process failure mode and effects analysis. These seven is the action validation step where we verify the implemented actions and attempt to reproduce the defect to ensure actions are effective. And the final step, D eight is lessons learned and closure. This is the last step where we capture the lessons learned, conduct a closure audit to ensure everything is done properly, and finally, celebrate the team and close. This is the 80 structure that we will go through together. One more information in this point, you can see various 80s in different industries. Steps can slightly differ according to the company that you work in. It is quite normal. But in this course, we will follow these steps which are extremely comprehensive, detailed and well structured. 3. D5 Introduction to Root Cause Analysis: Root cause analysis is the core of a problem solving study, verifying the underlying reasons of the problem. Here, we will deep tie with special tools like fishbone analysis, faulty analysis, and five fis analysis. We will find both occurrence and non detection root cause of the problem. This distinguish is important because both are quite different points which caused the problem to come out. In the meanwhile, we will investigate the systemic root cause of the problem as well as the technical one. Now let's take a look why we are using this fishbone FDA and fi five and let's understand how they are showing us, finding the root cause. The first step to start the root cause analysis is fishbone. Here we will make it brainstorming to identify potential causes. I am saying potential because those causes are not verified yet here. We will just write down every idea that comes from the team later then we will move those to FDA analysis. Here, we will verify whether they are really the contributor of the problem or not. We'll compare the good and bad status to see the exact differences. And after that, finally, we will move those verified case to five fives root cause analysis. And here we will find the root cause by asking why. This is the method to find the root cause, and we will do this for both occurrence and non detection. 4. Fishbone Diagram: Fishbone analysis, also known as Ishikawa, is a brainstorming technique used to explore potential causes of a problem. It consists of several fundamental categories. There are people, machine, material, method, and environments. This is the fishbone diagram. At the head, we write the problem and then we raise potential causes from brainstorming and place them under the relevant headers. The reason behind these categories, method, material, machine, people, and environment, is that they fundamentally cover every step of the manufacturing processes. However, this doesn't mean that we can't add more. If needed, we can include additional headers to make fishbone analysis more specialized. For example, you can add design if method is not a broad enough category or measurement specifically for non detection analysis. But for now, we will stick to the five. Now let's take a closer look at the headers in the fishpond diagram and what kind of potential cases we define under each. First, method. This refers to process standards. We basically question whether the standards are correct or not. Potential casos here may include incorrect procedures, unclear work instructions or ineffective process steps. Next is material. This covers anything related to raw material, components or the part itself. For example, issues like low quality materials, contamination or out of tolerance products can be identified. And the next is machine. This represents the equipment, tools or technology involved in the process. Potential causes under this category can include machine malfunctions, lack of maintenance, machine breakdowns or other equipment issues, and next is people in the other word man. This refers to human related factors such as lack of training, operator errors, miscommunication or workload issues. And lost is environment. This includes external conditions that may impact the process. Factors like temperature change, humidity, lighting, noise, vibrations, as well as the process tidiness and FIFS related issues can all affect outcomes. By organizing potential causes under these categories, we enjoy a structured and logical approach to identifying the root cause of a problem in the next steps. Now let's talk about how we use this diagram. In the fishbone diagram, we first define the problem at the head. Then based on that problem, we write potential causes under the relevant has. Let's take an example. The problem is that, let's say, whole diameter is out of tolerance. To address this, we will identify potential causes through brainstorming. Now let's bring in some examples. Incorrect cutting parameters under method, lack of training under people, lack of instruction, material deformation, and more. Basically, we are just throwing out ideas and listing potential causes that jure nothing is missed. At this stage, we don't yet know if the issue is caused by material deformation, machine problems, operator training, or maybe environmental factors. We simply write down every idea without judgment. Best result of the fishbone analysis is the one that is done with sufficient people, not crowded but less as well, because we need to hear all perspectives to ensure nothing is overlooked. A more important question is also, how do we guide the team effectively in this stage? First, if you are leading the problem solving team, especially for the root cause investigation, you should familiarize yourself with the process, especially if you don't work directly on the shop floor because understanding the problem details and having even a basic grasp of the process will help you engage the team and encourage them to share their ideas. Most importantly, we should avoid judging ideas at this stage and ensure others don't judge either. Because if people feel their ideas are being dismissed or criticized, they may hesitate to contribute and we don't want that. Instead, we should create an open environment where everyone feels comfortable sharing their thoughts, feedback because in this phase, there is no bad idea. So if we go back to the fishbone diagram, the next step after identifying the potential causes is to evaluate them. You might ask, shouldn't we avoid judging the content as we just discussed? Yes, during brainstorming, we shouldn't judge because we want to capture every idea from the team. But once that phase is over, before moving to the verification step in FDA, we need to filter them. This is important because otherwise we will have to verify every single item in the next step faulty analysis, which could lead to unnecessary time and even costs. The best way to filter this cause is through voting, a simple voting process. You can ask the team, what do you think about the temperature variations? If the majority agrees that it is not relevant, eliminate it. Don't waste time on it and repeat this for all the items that you have. And after this elimination process, the remaining ones become all filtered potential causes. These are the ones that to be moved to FDA analysis. However, if there are any doubts during this elimination, it is important to keep those items and move them to the next step. The goal of this elimination is simply to remove unnecessary items to avoid wasting time. But if you have doubts, you should insist on keeping the item. 5. Fault Tree Analysis (FTA): Fault analysis or FDA in acronym. In FDA analysis, we verify the potential causes identifying the Ichi Cava diagram to determine whether they are contributing factors or not. Our fundamentals are standard requirement, good part or status and bad part or status. The logic is to demonstrate the difference between good and bad by referencing the standard requirement. FTA analysis table simply consist of these key elements. The first and the second columns are for potential causes. We just move the potential causes from the fishbone analysis to the second column and we write the fishbone factors like machine, method, material, or others in the first column. In the fishbone analysis, we only identify the potential causes without verifying them. But here we will determine whether there are actual causes or not. The next column is the standard column where we identify what the standard required. For example, if the potential causes is lack of maintenance, which is under the machine factor in fishbone, let's say, then we need to check the relevant standard that defines that maintenance requirements. That could be a maintenance procedure or instruction. For the example, let's say, maintenance instruction and stating like laprocation must be done every 500 hours. At this stage, we first define the standard, which is maintenance instruction for this example and then state the requirement of that standard, which is lubrication every 500 hours. Next is the good part or status olum. We use both part and status because sometimes the root cause is related to the condition of a part and sometimes it's about the operational status. Here we write the good condition for comparison. However, a good part doesn't necessarily mean it meets all the standard requirements. It simply means this part didn't have our problem. If the problem is machine breakdown, we write the actual maintenance value of the time when the machine was working properly as we look for the good part data. For this line example, if the machine was operating well previously with abrication every 500 hours, we can refer that maintenance record as the good part. We capture abrication done every 500 hours. The good part, which we didn't see the issued problem is the same with the standard requirement. Next is we have the bad part or status coolum. Here we identify the bad condition where the problem actually occurred. If the issue is machine breakdown, we check what the abrication interval was at the time of failure. Let's say there was no labrication for 1,000 hours. This simply states the situation at the time of failure. At the time we produce not okay parts, we hadn't done labrication for 1,000 hours, but at the time we produced good parts, abrication had been done every 500 hours, we simply wrote down the actual status for both good and bad. At this point, we have both good and bad conditions documented. The good part complies with the standard and the bad part doesn't comply with the standard. And now we will assess this situation with four simple questions. First, we need to ensure whether the standard is correct or not because sometimes the standard may have lack of information or even be completely wrong. For example, if the standard requirement didn't specify the frequency of lubrication for this example, we will say the standard is not okay, so the answer would be no. But in this example, lblcation is every 500 hours is clearly identify, so we confirm that is okay, the answer will be yes. And the next is is the good part okay with the standard. The standard requirement states the abrication must be done every 500 hours and the good parts condition is labrication already done every 500 hours. Since this match, the answer is yes. The next is the bad part okay with the standard? The requirement is lubrication must be done every 500 hours as per the maintenance instruction. But the bad parts condition is no lubrication for 1,000 hours. This doesn't comply with the standard, the answer will be no. The final question is, is there a direct link between this cause and the problem? Here we simply evaluate the conditions by considering all the answers we gave for the standard and the good and the bad part statuses. The previous three answers of yes and no determine this final result. Let's continue from this example to understand better. In this example, we confirm the standard is okay and the good part follows the standard, but the bad part doesn't follow the standard. This difference shows there's a link with the problem, the answer will be yes. Because it is obvious that for this potential cuss, we produce a good part that complies with the standards. But when we check the Nooky parts production data, we see that. At that time, we didn't comply with the standard. This is very clear difference, which means we will make a further investigation into this case in the five Fs root cause analysis. However, before going to five fives analysis, it is always useful to try to reproduce the defect by using the same deviation in this specific cows line item. It is to ensure that this factor is directly causing the problem. For example, we should keep machine not lubricated for 1,000 hours like in the bed part status, and after that, if we see the problem comes out, then we can absolutely ensure that it is directly linked with the problem. While this method is very useful and recommended, where it is possible, it is not always possible at this stage. If the analyzed item is a technical thing that can be simulated quickly, it is very good and we can try to reproduce quickly to ensure. However, for example, for this labrication line item example, we should wait for 1,000 hours without applying any labrication just to simulate this failure mode, which is not possible and not make sense. As long as it is possible, we should try to reproduce the defect to demonstrate direct linking. If the final answer is yes, this means the potential cause is no longer just potential. It is a verified cause. It must be investigated further using the F fives method in the next step. But if both the good and bad parts comply with the standard, then the answer will be no. So this potential causes is eliminated as this doesn't contribute to the problem. There are a few different possibilities here that we need to consider. Let's go through a slightly different example and understand different possibilities. At this time, the potential cows and the standard are same, but good and bad part status are different. The good part is no lubrication for 1,000 hours and the bad part is lubrication every 500 hours. Here, the good part doesn't comply with the standard, the good part answer is no. But the bad part complies also with the standard. The second answer is yes. What about the last questions? Direct link to the problem. Even though the good part deviates from the standard, the analysis shows that lack of maintenance is not the cause for this line item. There must be another factor leading to machine breakdown because if the lack of maintenance were the real cause, then the bad part should also fail to meet the standard. But here, we don't have the situation, meaning this potential cause is not the real cause of the problem, there is no need to move to further investigation in FI FIS as part of this problem solving study. However, even though this deviation is not related to our current issue, we should still fix it to ensure that compliance. It is not relevant to our problem, but it is a potential concern and could cause other issues in the future. We should also fix this out of this eight study. Let's go through one different example with another possibility. But the good part and the bad part don't comply with the standards, no lubrication for 1,000 hours. So this point is a bit tricky because it is obvious that we produce good parts with this deviation and we hadn't done lubrication for 1,000 hours, which is out of the requirement and we didn't have an issue. On the other hand, the same situation happened when we produced the bad parts. So it seems this is not the cause of the problem. However, this might be one of the contributors. For example, when it combines with some other causes, the problem can occur. This indicates that lack of lubrication alone is not a direct cause even though it deviates from the standard, but it may be a contributing factor when combined with other conditions led to the failure. So for the answer to direct link, we can put a question mark in this example because we are not sure yet whether it is a direct cause or not. However, we should still investigate further in the 55 analysis, even though the cause is not verified yet because the obvious fact here is that both good and bad part don't comply with the standards, this might be a potential contributor to the problem. Now we have completed the fault tree analysis and we will continue with the last step of the root cause analysis, which is 55. 6. 5 Whys Root Cause Analysis: Five fs analysis is the final step in identifying the root cause. We started by identifying the potential causes using the fishbone diagram, and then we moved those potential causes to FDA analysis for verification. Finally, we take the verified causes into the five fs analysis to determine the true root cause. F fs analysis is a straightforward, step by step investigation. First, we write the verified case which come from the FDA analysis. Then we start asking the question of why repeatedly until we find the root cause. The key is that each answer must be data based, not subjective or estimated and each why answer should take us one step closer to the root cause. This deep investigation method is quite powerful as well as the straightforward. Let's go through a simple example to better understand this. Let's say we have a problem of machine breakdown and we created a fishbone diagram and then made an FDA analysis and found the real cause of the problem, which is motor overheated. We know that. This is the cause of the problem, but we don't know yet why did motor overheat. We will learn this with the 55 root cause analysis. Let's start. The first question is why did motor overheat? The answer is because the coolant system was not working properly. Then why wasn't the cooling system working properly? Because the coolant level was too low, why was the coolant level too low? Because there was a leak in the coolant hose. Why was there a leak in the coolant hose? Because the hose was worn out and not replaced in time. And then why was the hose not replaced in time? Because there was no preventive maintenance plan for the checking the coolant system. This is our root cause. What we have done is proceeding step by step and each answer is databased. We got closer to the root cause with each answer and finally, we found it. It is a lack of maintenance plan for the coolant system. To ensure we answered correctly, we can cross check by asking so to the backward. This is a verification. Let's do this. There was no preventive maintenance plan for the coolant system, so that the worn out hose was not replaced in time, so that there was a leak in the coolant hose, so that coolant level was too low, so that the coolant system wasn't working properly, and so that motor overheated. We basically verified our answers and it definitely looks correct according to our example. One other point here is sometimes we may have two or even more different sub answers after a question. This means you can have two or more root causes for one verified cause. Let's quickly go through an example to see how it works. For this example, let's use the same scenario. We started with the root cause of the verified factor, which was the motor overheating and the first three answers are the same. Why did motor overheat Because the cooling system was not working properly, and why? Because the coolant level was too low and why? Because there was a leak in the coolant hose. Now let's imagine that. The third question, why there was a leak in the coolant hose has two different answers leading us to two different root causes. The first answer is that the one we already identified, lack of a maintenance plan. But we also have a second root cause and it is about low material quality of the horse. We asked the same question, why there was a leak in the coolant hose, but identify two answers. The first one is the hose worn out due to lack of maintenance plan, and the second one is the horse material was low quality. Both of these are varied root causes of the motor overheating issue. The root causes we identify here are technical root causes which are common to all problem solving studies. However, we have also one more type of root cause, which is systemic root cause. This is generally not done for every problem solving study and not a mandatory requirement, but it undercovers all the underlying systemic causes and helps you build a robust system. Let's move on to the next lecture and see how we do this. 7. Systemic Root Cause: Systemic root cause is the cause of the technical root cause. Basically, we just ask one more why to identify this. This final answer will lead us to lack of standardization, which is the systemic root cause. Let's take the previous root cause example. The root cause that we have identified was no preventive maintenance plan for checking the coolant system. Now let's go one step further and ask one more why. Why there was no preventive maintenance plan for checking the coolant system. The answer is because there is no structural procedure and strategy for maintenance. In the other word, lack of structured maintenance procedure. That is the reason for the lack of maintenance for checking the coolant system. Here, this final answer is our systemic root cause, which highlights the lack of standardization and the systemic factor behind the issue. While systemic root cause analysis is not commonly applied in the industry, it is very useful to include this because this approach allows you to implement preventive actions easier in the next phase. This concludes the root cause analysis, so now we will continue with the case study. 8. Explore the Case Study: We will proceed with a case study for all the steps after the theocal lectures. We will use the same case study that we went through in the previous course, D one and D four. But let's do a quick recap for those who haven't taken the previous course. We are a bonnet hinge manufacturing company and our customer is a car manufacturer. Basically, we produce bonnet hints and ship those to them. We received a problem from the customer. The issue was a misalignment in the second hull of the bonnet hinge. Due to this misalignment, the customer was unable to assemble the hinge onto their cars. Basically, the whole position on the hinge was incorrect. Let's bring the visuals of the bad and good part to see this difference clearly. This hinge consists of two components. The one with the three hose is not okay due to the position of the second hull. Now let's go through our production process of the bonotinge. First, we receive sheet metals from our supplier which are then stored in the receiving material warehouse. Next, these sheet metals are fed into the press operations. In the first press operation, we produce the first sempduct and these semi products are then stored in the intermediate storage area located by the production line. After that, we produce a second semi product, which like the first is stored in the intermediate stock area by the line. In the final operation, we assemble these two sempducts to produce the complete bono hinge. The finished products are initially stored in the intermediate stock area and then move to the shipment warehouse where they await for dispatch. From there, we ship the complete bono things to our lovely customer. In summary, our production process involves creating two semi products that are assembled to form the final product, the bono hin. Now, looking at our issue, the faulty component is produced in the second press operation. Therefore, our focus in the production process will be here. The second press operation is the key area that we need to investigate, as this is where the failure occurred. 9. Case Study - Fishbone for Occurrence: In this lecture, we will start the root cause analysis for the occurrence through the as study using the fishpond diagram, which is our first tool. We will identify the potential causes for the occurrence, which is the first part of the root cause analysis. Now let's start the fishbone analysis. The problem was bonnet hinge second hole mislocated. In this study, we will investigate the potential causes of the issue for the occurrence. We will make a brainstorming session to identify the possible reasons for the problem specifically by focusing on occurrence reasons. We should try to have expert people in the team during this study, but at the same time, some fresh minds can bring different perspectives, a mix is always beneficial as much as possible, of course. During this brainstorming with the team, we don't interrupt or judge any idea. We just listen and write exactly what is said to encourage participation. And once we identify all the potential causes, we will make a pre elimination to avoid wasting time in the next step FDA. Now let's start identifying all potential causes with the team. We have a lot of different potential causes. Some of them may be quite irrelevant, but others can be directly related to the problem. Now that we finished writing all, let's make a pre elimination with the team. This can be done through an open discussion or a voting system. But the best method is to convince as many people as possible in the room. Let's start from the method. Insufficient tool setup checklist. The problem is that a specific hole is consistently in the wrong position over a period of time. So this suggests that the tool setup was incorrect from the beginning, since the part cannot produce incorrectly unless the tool itself is misaligned. Insufficient tool setup checklist could be a contributing factor. So therefore, we don't eliminate this as we want to keep it for the next level analysis. It can be a factor in the problem. Next is lack of change management. In this case, a specific hole was consistently misaligned. This potential case suggests that the tooling setup was altered at some point and not properly controlled. If any modifications were made to the tooling, they should have been properly documented, verified, and reverted if necessary. But something could have gone wrong if this is relevant. It seems highly relevant to the failure mode and requires further analysis in FDA. Next is no tool maintenance plan. The absence of a tool maintenance plan that includes tool conditions checks could allow such issues to persist over time. Regular maintenance should include tool conditions, it is relevant. We can absolutely review this potential factor in FDA. Now let's continue from the material. First is sheet metal elasticity. The variations in sheet metal elasticity could cause warping or deformation. But this issue specifically relates to the whole misalignment, we eliminate this as it is not relevant to the problem. Next is sheet metal thickness. If the material thickness were inconsistent, it could lead to general issues with the whole shape or maybe depth. However, since misalignment is occurring in a specific direction, this is not relevant ta. Last one in the material section is material contamination. Material contamination such as dirt or foreign particles can affect the quality of the part, but it typically results in surface defects or imperfections, maybe. However, there will be no impact on the consistent whole position. No need to make further analysis for this. Now let's continue from the machine. First is incorrect setting of machine parameters. The issue is that a specific hole was consistently wrong position. This suggests that tooling misalignment rather than an incorrect machine settings. If machine parameters such as pressure, speed or stroke length were incorrect, we would like to see defects like inconsistent hole positions, deformations, or improper cutting rather than a consistently misaligned hole. It seems that we can eliminate this potential factor. Next is insufficient press tonnage capacity. The issue is consistent hole misalignment, not general tool deformation or insufficient press force. If the press had insufficient tonnage, we will likely see a general part deformation, not just a misaligned hole. T us, this is not relevant to the issue. Next is lack of maintenance. The issue is misalignment of the punch, not due to wear or damage from the lack of maintenance. If maintenance for the cause, we will see more widespread tool issues or general wear, not a specific misalignment. Let's eliminate this one as well. Last is inconsistent press calibration. If the press calibration were inconsistent, it could cause slight variations in the punch alignment during the production, even if the tool setup was correct. But the point is our second hole is misaligned, not the other holes. If calibration were the issue, all the holes will be affected, so we can eliminate this as well. Now we move on to the people in the other word man. It is inadequate training on punch alignment. If the relevant operators lack training on proper punch alignment, they may fail to notice misalignment during the setup, so this could lead such issues. It is worth investigating further in FDA. Next is environmental factors. We have just one factor here, which is poor lighting in the setup area. Poor lighting is not a strong cause for incorrect adjustment of such large tools. Therefore, we can eliminate this potential cause as well. So now we have identified all potential causes for the occurrence with fishbone analysis. Now we will make further analysis in FDA for total four different potential causes from here. Let's go to the next lecture and see how we do this analysis. 10. Case Study - FTA for Occurrence: This is our FTA table, fault analysis. Now we will bring the potential causes that we have identified in fishbone analysis and analyze whether they are really a cause or not. Let's start with the first one insufficient to setup checklist. This comes from the method category in fishbone. Here, we will evaluate whether the checklist was sufficient to prevent this issue The standard and the requirement here is that the two setup checklist must include punch alignment verification to ensure correct hole placement. Now let's check the good part or status. During the production of good parts, was this checklist used and did it include punch alignment? The answer is yes. The checklist included punch alignment verification. But what about the bad parts? Were we using the same checklist when we producing the not okay parts? The answer is also yes. The checklist included punch alignment verification. So now let's evaluate what this means. First, we check the standard itself. The requirement is the two setup checklist must include punch alignment verification, and we confirm that. This is okay. So the answer is yes. And next, we check whether the good part complies with the standard. The answer is yes, again because it meets the requirement that we defined. And then we check the bad part. The answer is, again, yes, because the checklist included this requirement even during bad part production. So at this point, we see that there is no difference between the bad and good parts for this specific line item, but seem okay. That means there is no direct link between this factor and the problem. Since these potential causes doesn't contribute to the issue, we just eliminate it from further analysis. So there is no need to move this to five fs analysis. The next one is lack of change management. In this case, there is no documented standard, but the expectation is that any temporary change in tool setup must be properly documented and reverted before production resumes. Let's check the good part or status. To do this, we will review the production or tool revision history. Here we see that no modifications were made to the tool since production run started. Production followed the standard setup. But in the bad part, the punch position was modified for a trial but wasn't reverted back afterward. Which led to continuous misalignment in production after the trial. Clearly, a test trial was conducted, but it wasn't managed properly. Now let's check the standard. The expectation is that any temporary tool change must be properly documented and reverted before production resumes, which makes sense. However, we don't have the standard documented anywhere. It is just an expectation for this type of modification activities. We confirmed that the standard is not okay because it is not written anywhere. It is just an expectation. So for the next step, we will consider their status by treating the expectation as a standard. In that case, the good part complies with the expectation. So the answer is yes. But the bad part doesn't comply with the expectation. So the answer is no. Now, is there a direct link between these cause and the problem? Yes, because the bad part doesn't comply while the good part does comply according to the expectation. On top of that, we don't even have the standard documented anywhere. So this lack of change management absolutely needs to be investigated by Fis root cause analysis. Our next potential cause for fish Bon is no tool maintenance plan. The standard requirement is that a maintenance plan must be in place, and it should include tool condition checks to ensure long term stability. Now, let's check the good part or status. Here, we need to verify whether the maintenance plan existed and was valid during the good part production. Based on the maintenance records, we see that the maintenance plan exists and included the tool condition checks, meaning that process follow the defined structure. Now let's check the bad part. When the failure occurred, we again see that the maintenance plan exists and included tool condition checks. The maintenance process itself was not missing and the same conditions apply to both good and bad part production. Now let's check the final yes and no questions. The requirement states that the preventive maintenance plan must include two condition checks. In this case, we confirm that such a plan exists and documented. The standard is okay. That means our answer is yes. For the next questions, both good and bad parts are under the same maintenance conditions. They comply with the standard. The next two answers are also yes. There is no direct link because both good and bad parts were under the same maintenance plan. There was no difference between them regarding this factor. Therefore, this factor is eliminated from further investigation. Next is inadequate training on punch alignment. Here we are checking whether the operators received proper training on how to verify punch alignment during setup. Let's start with the standard. The requirement is that operators must receive training on punch alignment and during, they can correctly verify its positioning before production. In this case, the standard exists and defines the training requirement. We confirm that the standard is okay at the moment. Now let's check the good part or status. During good part production, the operator had received training, but training didn't sufficiently emphasize punch alignment verification. Now let's check the bad part. Here we see the same situation. The operator was trained, but the training lacked enough focus on punch alignment. So training conditions were identical for both good and bad part production. Since the standard is okay, but neither the good nor the bad part comply with the standard. So both answers are no. So the last question, direct link with the problem, we are not completely sure yet because both answers are no, we cannot confirm it's a direct cause, but it might be a strong contributor factor maybe. So we mark it with question mark and carry it forward to F five's analysis for further investigation. Now we have completed verification of four different potential causes in the fault analysis. We confirmed two of them as actual causes of the problem. Now we will move them to F F root cause analysis, which is the final step of the root cause investigation. Let's move on to the next lecture and see how we conduct FF root cause analysis for these causes. 11. Case Study - 5 Whys for Occurrence: So now we are going to conduct a 55 root cause analysis for the causes we have verified in the previous FTA analysis. Our first verified cause is lack of change management. So let's start by asking the question of why and diving into the underlying reasons behind this lack of change management. First, let's start with the question. Why was there a lack of change management? So instead of starting with a problem, which is the whole misalignment, we directly ask that question because this case has already been verified in the FDA analysis. Our answer is because there was no documented rollback process after temporary tool change. Now let's move to second question. Why was there no documented rollback process? The answer is because the change management procedure for temporary tool adjustment was not clearly defined. Here we already identify the technical root cause. So our management procedure was unclear when it came to the temporary tool adjustments. So this is the technical root cause of the problem. That means if we had a clear procedure in place for managing temporary tool adjustments, we wouldn't have had this whole misalignment problem. Now let's take it to one step further and ask one more why to explore the systemic root cause. Why were the change management procedure for temporary tool adjustments not clearly defined? The answer is because the organization lacked the awareness to anticipate the risks with temporary tool change. This is the systemic root cause of the problem. So by starting directly from the verified cause, we have shortened the five fis process, and that's completely fine because our aim here is to conduct a precise analysis to get to the real root cause. However, in some cases, some customers may ask for a detailed step by step breakdown of the five fis process. And in those cases, you can directly start with the problem instead of the cause. So for example, why was the second hole mislocated Be second punch position was not correct in the tool, and why? Because it was modified by the tool maintenance team for a trial and not reverted back afterward. And why? Because there was no documented rollback process after temporal change. The result is same. We just extended the 55 process. Let's continue with the other casts from FDA, which is inadequate training on punch alignment. Why was the training on punch alignment inadequate? The answer is because the training content didn't sufficiently cover punch alignment verification during the setup. Now let's move to second question. Why did the training content not sufficiently cover punch alignment verification? Because the training scope was not clearly defined to ensure all standard requirements were fully covered. This is the technical root cause of the problem. If the training scope was prepared by covering all the requirements, punch alignment would have been clearly addressed during that training. But we can still ask one more question to understand the systemic root cause under it. Why was the training scope not clearly defined to ensure all standard requirements were covered? Because there was no structured qualification or competency requirement for those preparing the training content. This is the systemic root cause of the problem. In the technical root cause, we address that specific training content and if we correct this, we would have eliminated the issue for this specific problem. But now we know one more issue. We don't have a structured qualification system for those who prepare training content and this gap can cause different problems in different process and different parts because this is a systemic root cause. Now we can also verify the logic of this root cause analysis by asking so basically a cross check. Let's do it together. We will start from the end. There was no structure qualification or competency requirement for those preparing training content. So that training scope was not clearly defined, and so that training content didn't sufficiently cover punch alignment verification. You see the logic, right? Instead of asking why in a straightway, we just ask so starting from the end for a cross check. One another point in this analysis, you might ask the issue origin was change management since they didn't revert back to the punch after the trial. Why do we have a root cause here for punch alignment training? Well, one problem might have multiple root cause. For example, in this case, lack of training could be a contributing factor because the tool training was not covering punch alignment. We detected this gap in the FDA analysis and verified that both the okay and not okay conditions in FDA show this deficiency. Therefore, this must absolutely be addressed and be resolved. This concludes the root cause occurrence case study. Now let's move on to the next and analyze non detection. 12. Case Study - Fishbone for Non-Detection: For the non detection, we will start with fishbone again by identifying the potential causes which are relevant to controlling of the part in that time. Basically, we will focus why we didn't detect the failure part. Let's start the brainstorming and identify all potential causes raised by the team. First is no control for hil position. This should be checked in FDA because it is directly relevant to the failure mode. Our issue was the misalignment of the second hole on the bonnet hinge. If there was no proper control in place or if the existing controls were not effective enough, this could have contributed to the problem. As a team, we identified this as a risk factor and decided to investigate it further in the fault tree analysis. Next is insufficient control to detect the failure mode. This is also an important point because if the current control methods were not sufficient to detect the misalignment, we need to understand why. Maybe the inspection method was not capable or the frequency of checks was too low. Since this directly relates to the non detection aspect of the issue, we are taking it to the FDA for further analysis. Now let's move on to the machine category. First is lack of MSA on control fixtures. MSA is measurement system analysis. If the measurement system itself was not capable, then even if a check was done, it might not have been effective, we decide to take this to FDA as it can be relevant. Next is lack of calibration of control fixtures. Calibration is necessary to maintain measurement accuracy over time. If the control fixture was not properly calibrated, it could have led to false acceptances, meaning misalignment holes might have passed inspection. Therefore, this is a potential contributing factor to the non detection, and we mark it with cross to further analysis in FDA. Last one in the machine section is assembly fixture deformation. The assembly fixture is also used for enduring proper alignment during the assembly process. If any deformation or mole function occurs in the assembly fixture, it can lead to improper alignment of the parts, making it impossible to detect the issues. Therefore, this might be relevant to potential contributor for the non detection. Let's mark it for further analysis in FDA. Now let's move on to the people category, in other words, man. First is lack of training. Training gaps can lead to improper execution of control. For example, if operators weren't trained well enough, they might not have followed the correct procedures. This could be a contributing factor and it should be analyzed in FDA. And second is inadequate supervision. Spervision for the controls is not a requirement and expectation as all responsibilities were already clearly outlined in the responsibility matrix. So therefore, as a team, we eliminate this potential cause. And the last one is FFS is not proper under environment. This is also another one that we eliminate rather than moving to FDA because the problem is not singular and it is quite unlikely to influence the control mechanism in such kind of non singular problems. So that's why we eliminate this. Now let's move to FDA for verification. 13. Case Study - FTA for Non-Detection: Now we bring the potential causes from fishbone analysis to verify in this FDA table. Let's start. First is no control for whole positions. It is from method in fishbone diagram. Here we check whether there's a control for the whole position or not. We will not check the control is done because this line is not about that. I just questions whether we have control mechanism in place or not. Let's go through the questions in table. Process standard and requirement. It is PFMEA, process failure mode and effects analysis. There are two controls. First is five parts every hour at press production and second is 100% frequency at assemble fixture. Regarding good parts, control plan has the controls, and regarding the bad parts, control plan has the controls as well, no difference. So that standard is okay. Good part is okay according to the standard and bad part also okay according to standard. That means there is no direct link. Problem is not about this potential cast, so we eliminate this. Next is insufficient control to detect the failure mode. Again, it's coming from method. Here we will check whether the current controls are enough to detect the whole misalignment failure or not. Again, here, we will not check the control is done or not. We will check the current controls capability to ensure they are enough. Our requirement for these potential causes is current controls must detect the whole misalignment. This is the requirement that we expect from the current controls. Regarding good part and status, current controls detect the whole misalignment. They are capable. And same for the bad part because we had same controls in the control plan. So the standard is okay. Good and bad parts are comply with the standard. Therefore, there is no direct link with the problem and we eliminate these potential cows. Next is lack of MSA on control fixture. MSA stands for measurement system analysis, which is the methodology to verify the control systems. It is different from calibration. Standard is MSA must be done every year and result of MSA must be okay. When we check the status of MSA for both good and bad parts, it seems okay. MSA done just one month before the problem happened, it is vd. There is no problem for MSA. Therefore, standard is okay, good and bad parts are comply with the standards, we say yes for both them and there is no direct link with the problem. This one eliminated. Next is lack of calibration of control fixtures. Here we will check the calibration status on the time period, both good and bad parts production. Standard is control fixture calibration must be done yearly. Let's check the good part. What we had in that time, control fixture calibration was vet when we produce good parts. Regarding bad parts, it is control fixture calibration wasn't vet when we produced bad parts. It clearly shows that there is a problem here. Let's bring that control fixture to understand it better. This is the control fixture and it controls three holes of that part in this way. But the problem is, it doesn't have a valued calibration at that time. So let's continue the other questions. The standard is okay, so we put yes. Good part is okay according to the standard. So this is also yes, but the bad part is not comply with the standard. So that is no, and the result is there is a direct link with the problem. So we will move this to fi five Root cause analysis for further investigation. Next is assemble fixture deformation. It is formation. The requirement is no deformation on assembly fixture. However, when we check the assemble fixture, we see that second pin is missing, which is the control point for the whole existence. The requirement is no deformation on assemble fiixture and the good part is no deformation and assembly fixture and all control pins are in place. This shows that there was no issue when we produced the good parts, which were okay for whole alignments. Let's check the bad part. Second pin is missing. When we produce the bad part, second pin is assemble fixture is missing. Let's bring that assemble fixture to better understand. This is the assembly fixture. Two semi products are placed and assembled in that time. Three holes are being controlled by those three pins. That means if the parties assemble it, then holes exists. But if you see a bad situation, the second pin of the fixture is missing. Therefore, it couldn't detect defect during this assembly operation. Let's continue to other questions. Standard is okay. The answer is yes, good part is comply with the standard, so it is yes as well, but the bad part is not comply with the standard because second pin is missing. The answer is. Therefore, there is a direct link here which needs to be analyzed in Fifi. Nexus from the people, in other words, human error. It is lack of training. Here we will check the control operators have the relevant trainings or not. Our standard requirement is operators must be trained for the control fixture and assembly fixture checks. This is the requirement according to that potential cause. The good part is trained operators work at that time. Let's check the bad part. It is same again. Operators had the training when bad parts were produced. Standard is okay. Good part and bad parts are also okay according to standard, there is no direct link about the lack of training. This one eliminated. This was the end of the non detection FDA analysis. Now we will move these two verified case from the non detection into the 55 root cause analysis and we'll make a further investigation. 14. Case Study - 5 Whys for Non-Detection: So now we are going to conduct the 55 root cause analysis for the non detection. Our first verified case from FTs lack of calibration. Let's start by asking why. Why was the control fixture not calibrated? The answer is because it was sent to calibration and not returned in time. Now as you see, the root cause forks to two ways. That means we have two different causes of this. So let's proceed at the left side. Why was the production continued while the fixture was still in calibration? The answer is because there was no clear rule preventing production from running without the fixture. Why? Because process controls didn't include a mandatory check for fixture availability before production. Why? Because the production startup approval process hasn't got the full equipment list. This is our root cause because obviously, if we had a full equipment list in the production startup approval process, this control would have done in a structural way so that we could detect this. Now let's ask one more why to go a bit more deep to understand the systemic issue underlying of it. Why was the equipment list not properly considered in the production startup approval? The answer is because there was no formal procedure to ensure all required equipment was checked before production started. This is the systemic root cause behind that technical one because it shows the lack of standard approach and once we solve that, the technical root cause will not be repeat in the next time. We want Skip to put full equipment list in the production startup approval form because we will have a formal procedure to describe how to create these control forms. Now let's continue from the right side, which is going through a different way. First cause was the fixture was sent for calibration and not return in time. The next question of why should be why there was no alternative control method. The answer is because no backup inspection method was defined and why because no contingency plan in place for the alternative control options like CMM or other manual measurement methods. So this is the second root cause of the problem. Second technical root cause. So let's ask one more question to see underlying reason of this lack of contingency plan. Why we don't have a contingency plan for alternative control options? Because there is no document requirement to establish alternative inspection methods. This is the underlying cause, which is the systemic reason. So if we solve this, that means we will prevent the risk of no contingency plan, which is the technical root cause. Let's continue with the other verified cuts from the FDA, which is assembly fixture deformation. So let's ask why why did the assembly fixture deform? The answer is because second pin was missing, which is for checking the alignment of the oil, and why? Because the second pin was remote for maintenance and was not reinstalled, and why? Because there was no formal maintenance step to ensure all parts of the fixture were checked and reinstalled correctly after maintenance. And why because maintenance instruction of the equipment didn't include a step for confirming that all fixture components were checked and reassembled properly. This is our root cause, the technical root cause of the assemble fixture deformation. Let's ask one more why to see lack of main standard. Why didn't the maintenance instruction include this verification step? The answer is because the maintenance procedure didn't emphasize the importance of enduring the fixture functionality and condition before it was used in production. This is the systemic root cause. We chose the lack of main standard. This was the end of the root cause for the case study. Now we will quickly summarize what we have done in this whole section. 15. Summarize D5 Root Cause Analysis: Summarize the root cause analysis and key takeaways. At first, we started with the Ichikawa, in the other words, fishbone analysis, which led us to identify potential causes of the problem. We focus on the problem and throw the ideas which can cause the problem. These ideas were the potential causes and after a quick check, quick filtration with the team, we move these to fault three analysis, FDA. Here, we verify those with a few questions by comparing the good and bad parts. In this analysis, some of them verified, but some of them just eliminate it because they were not relevant with the problem. Those verified cause moved to five fives annals and investigated deeply, which is the final step of the root cause. We have found both technical and systemic root cause in five fives. While technical root cause was the origin of the problem, systemic was more about the lack of standardization, which affects more area. Distinguish is important because in the next step during the action plan, we will link the corrective actions with the technical root cause and also we will link the preventive actions with the systemic root cause. Additionally, we touched the verification of the root cause, which is important to ensure we answered correctly. It was like a cross check by asking so starting from the end to the backwards. This was the end of the root cause analysis and now let's make a quick quiz to reinforce our learning. 16. D6 Introduction to Corrective & Preventive Actions: Corrective and preventive actions. In this step, we will understand the difference of these two types of action and learn how to create a robust action plan. In the previous step D five, we identify the root cause both for occurrence and non detection. Now here, our aim is to define and implement action against these root cause. In a simple say, eliminate the root causes. After that, we will review and update PFMEA, process failure mode and effects analysis. We have occurrence and non detection root causes. Underneath, there are technical and systemic root cause. Again, both for occurrence and non detection. Here in action plan, we will fix the technical root cause with the corrective actions and we will fix the systemic root cause with the preventive actions. The planning of these actions will be in a simple format like action, responsible, due date, and action date. Now let's move on to the next lecture and see how we will create this action plan. 17. Corrective and Preventive Actions (CAPA): Corrective and preventive actions. In this step, we will understand the difference of these two types of action and we learn how to create a robust action plan. In the previous step D five, we identify the root cause both for occurrence and non detection. Now here, our aim is to define and implement actions against the root causes. In a simple say, eliminate the root causes. Corrective and preventive actions can be identified in a simple timing plan format. At first column, we have reference number, then the root cause column to link the root cause with the action. Next is the action type, whether it is corrective or preventive. Then the action itself, that means what we will do against that root cause. After that, responsible, who will conduct or lead the action and the due date for a clear timing, and finally, actual date of the action implementation. Now let's make an example to understand how it works. Let's say we have a problem of machine breakdown and our root cause for the occurrences, no preventive maintenance plan. Since this is a technical root cause, our action to fix it will be corrective. Simply, preventive maintenance plan will be created. Action responsible and the due date have also been identified like that, and once the action is done, then the actual date to be written in the last colum. So how do we know that the action we put here corrective or preventive? Well, the answer is simple. If you take action to fix the technical root cause, then it is corrective in this example, because you correct, you fix the root cause. But if your action fixes the systemic root cause, then it is preventive. Let's put here a systemic root cause of it. It is lack of structure maintenance strategy, which is the cause of that no preventive maintenance plan. Simply here, if we take action against this one, like a comprehensive procedure for maintenance activities will be created, this action will fix directly the systemic root cause. Therefore, it is preventive. Alternatively, you can also think like that. If you have one root cause and didn't deep dive for a systemic one, let's say, if your action is fixing that root cause, it is corrective. But if it prevents the reoccurrence of the root cause, then it is preventive. In this perspective, you can see this example. Creation of maintenance plan will fix the first root cause, which is the technical root cause. But when we create the procedure, it will prevent the reoccurrence of this first root cause. This is the logic. This distinction is simple but crucial to identify your action as corrective or preventive. This action plan is not about the problem itself. It's about the root causes that we identified because the problem should have been fixed in containment actions, either in temporary or a robust way. But here in corrective and preventive action plan, we focus on the root causes. In the other words, the underlying reasons of the problem itself. One other thing is we can have multiple actions for each root cause. For example, we can add one more action here as training will be organized about maintenance procedure, which is feeding the second action here because we should also think about the impact of our actions. If there are other affected standards with our current action, we need to also update this as well. And the last point regarding the timing of the actions, we should try to provide early dates as quality issues are always priority both for the customers and operations. While there is no global rule, in general, definition of the root cause and planning of the actions can be targeted as ten days and full eight decomplation can be set as 30 days. But as I said, it is completely depending on the customer that you work with and your own standards. 18. Understanding PFMEA (Extra session to understand the logic of PFMEA): FMEA, failure mode and effects analysis. This is a proactive tool that help us evaluate and analyze potential risk and take action to prevent failures before they happen. It's a structured and robust method to identify risks in production process, design, or even entire systems. There are different types of FMA, such as design FMEA, process FMA, or system FMA. In problem solving studies at manufacturing environment, we mostly use PFME because PFME is used to identify the process risks and define controls against us. Here in AT study, this is one of the key standard that we need to review and update according to the failure mode, root cause, and the actions that we have identified. In AT studies, we don't make a full FMEA study. We just review and update it where it is necessary as part of the AT study. But to do this, we need to know how it works. In this lecture, we will learn its whole logic. This is a process FMA template, which is Edge forth revision. Let's review these columns to understand their function. First column is process where we have the operation step. So if we proceed with a simple example, let's say, our process step is drilling operation. And second, is requirement. Here we identify the requirements of this operation. So what we expect from this drilling process? Let's say we expect a hole in ten millimeter plus and -0.05 millimeter diameter. Next question is potential failure mode. Here we ask what are the failure modes of this requirement. For this example, if we expect ten millimeter diameter with 0.05 tolerance, then it can be smaller or greater. These are the failure modes of this requirement. For this example, let's just say diameter is smaller. Diameter can be smaller than 9.95 millimeter, which is lower tolerance. This is a failure mode. Next is potential effects of this failure mode. Here we evaluate the potential effects of the failure. For this example, what does smaller diameter cause? Let's say, it causes an assembly problem in the process. Counterpart doesn't fit, impossible to assembly. This is the potential result of this failure mode. Next one is severity rating. Here we rate these potential effects of the failure mode out of ten. We have a table to do this, starting from one, which means no dissonable effect to ten, which means safety or regulation issues. For this example, this is about process effect. There is no customer effect for this failure mode because if they cannot assemble the part, it will not be even sent to the customer. Therefore, we will use the second coolum in this table. And we can rate eight, which is very high because there's a risk to stop production. Next to it, you see class olume. This is about the product or process special characteristic. If there is. For example, if you are familiar with the technical drawings, you can see some special characteristics such as CC or reverse Delta. This special characteristic mostly come from the product or function or maybe safety or maybe for regulation reasons. These symbols are changed depending on the organization and the customer. But for this example, this characteristic doesn't have any classification. Next is potential causes of the failure. Here we identify the underlying reasons of the potential failure mode. For this example, we said that this characteristic can be smaller than the requirement. That was the failure mode. We should consider what can cause that dimension is smaller than 9.95 millimeter, which is the lower tolerance of so let's identify drill bit were. So this is a drilling operation, and if the tool deform or wear in time, then hole can be smaller. By the way, there might be a lot of different causes that we can identify here, but we will go through in a simple way just to understand the logic. Next is occurrence, and just next to it, prevention control. So here, before rating to occurrence out of ten with the table, we first identify the prevention controls because we will consider both causes and the prevention controls when we rate occurrence. Let's identify the prevention controls. Here we think that what can we do to prevent drill beat wear? How can we prevent this or decrease the possibility of its occurrence? Prevention controls address the potential causes of the failure. Let's identify two measures here. First is production startup check to ensure drill bit is in good condition, and second, is changing the drill bit for every 4,000 parts. We have basically two preventions in place. Now we will rate the occurrence accordingly. That 0.1 of the important thing is constant the quality problems of other similar parts or processes if available. Because if you have a lot of problem in another similar product, although you have the same preventions, then the occurrence rate will be very high. But for this example, let's say we don't have much failure for the other similar parts. We just put three as it is a low possibility, which means one in 100,000 as estimated frequency. Next one is detection control. Here we assume that failure is occurred. Regardless of the occurrence possibility, it just occurred. How could we detect the failure mode or it's error, which is the cause of the failure mode. If we can directly detect the cause of the failure mode, which is drill bit with some automatic control, let's say, that's the best because you detect the error before it becomes a failure mode, but that is not always possible. If we cannot award or detect the error, then we need to detect the failure mode directly, which is the whole diameter is smaller than 9.95 millimeter for this example. How can we detect this? Checking as visual or putting some camera controls or having a control fixture. Let's identify a gauge control. Go and no go gauge. We will do this control for five parts per hour. The frequency is also important, but all other control details are identified in the control plan. For PFMEA, this level of detail might be enough. Now the detection rating. According to this control, what would be the detection score out of ten? We can rate as six because it is failure mode detection, post processing by operator through use of variable gauging or in station by operator through use of attributing gauging like go and nog gauge. It is 66 is the proper for attribute type of control, which is in station. Now the last is RPN risk priority number. Here we multiply these three scores, severity, occurrence, and detection, and the result is our RPN score, which is 144 for this example. This number is important to see the priorities. So for example, if you have three items and if one's RPN is 200 and the other is just 30, then you should focus on 200 RPN and make some more improvement, either in prevention or detection controls to reduce the number. These extra improvement actions are taken in the right section, which starts with recommended action. In case an improvement needed, we can identify new actions here and re evaluate the score again. For example, if we put a Pokao control, then the detection score decreased from 62 or maybe one according to the Pokao control type, then the RPN number would be reduced. Let's make one. Let's say recommended action is a camera control. Basically, camera will check the whole diameter automatically and if it is out of tolerance, it will lock the line which is a secure and robust control method. And next coolum responsibility and targeted are also identified like this and the next is action taken is blank because we didn't apply it. It will be done later. We just planned it at the time. NIF scores are severity is not changed, still eight, severity never change unless the part design or some requirements change because even you put a better control mechanism or even you put some prevention actions to avoid 100%, let's say, severity is still the same because it just considers the result when it happens. If it happens, the result is same, then the severity is same. No matter we put some Pocaok controls to eliminate to occurrence or perfect detection. Severity is always same, so it is eight again. Occurrence can be changed if you put better preventions. But in this example, it is three because camera control is about the detection, not about the prevention. Therefore, occurrence is not changed, it is three, but the detection is now three, as you see. We decrease it 6-3. According to the detection table, three is failure mode detection in station by automated controls that will detect discrepant part and automatically lock it, which is exactly matching with our nv camera control. Now the latest RPN is seven to two. But if it put some automatic control to detect directly the drill bit are, let's say, which is the cause of the failure mode, then we could rate a two because it prevents the failure mode by detecting the error itself. These kind of controls are generally called S pokaok. These are mostly rated as one or two in detection table. So now we did a simple example for the process failure mode and effects analysis. This is the main logic of P FMEA, and once the FMEA is completed, control plan is created accordingly, which includes directed controls like control type, control frequency, controller, whether it's a machine or operator, the characteristic to be controlled reaction plans in case failure detected et cetera as I said, in problem solving study, we don't set up a process. We just analyze a failure mode to resolve. So therefore, we don't make a whole PFME study or are control plan. We just review and update if necessary because at that time, it's already in production. So the necessary documents must already be in place. We just review and update them accordingly. Now let's move on to the next lecture and learn how we review and update PFMEA in AT studies. 19. Review PFMEA: Now we will understand how we review and update PFMA in AT studies. This PFM is from the same example but extended with different operations like chafing and packaging in addition to the drilling. First, the part is drilled, then chafing operation for the H, and finally is packaging. Now let's imagine we have a quality problem about contamination and we made some root cause analysis, identify some actions, and now we need to review PFMA part of this AT study. First thing that we need to do is finding the relevant failure mode here, relevant line item. Here it is contamination on the part, which is in the packaging operation. We have this failure mode. That means we have already evaluated the contamination risk. The only thing that we'll do updating this failure modes line item according to the new findings and actions. But if we didn't have this failure mode in PFMEA, let's say, we will need to create a new line item for the contamination failure mode. So for this case, the quality problem of the contamination caused by dirty working environment, the potential cause is not here as we just found it in the analysis. What we do is just adding it here. That's it. Regarding the current prevention controls, we say no control because here we show the status during the failure occurrence. All inpmpt action will be identified in the recommended action section in the right side. According to this data, we now rate the occurrence. Since we already have a failure for contamination, which was caused by dirty working environment, and since there is no prevention control for that, we rate high and put seven by using the occurrence ranking table. Regarding the detection control, it is same. We check visual. Therefore, detection rate is same with the above and the risk priority number is now 196. Now we will add the controls that come from the action plan. Let's say we adapt five a check at each startup to prevent dirtiness. We add this action in the recommended action section and we re evaluate the rating score. Severity is still same for occurrence reduced 7-3, which is low now because we improve the prevention by adding a five a control. And detection is still say we didn't change anything. It is seven. The RPA module which is obtained by multiplying of these three numbers is now 84. As we did this update just after the action plan, the action stays open because we didn't complete the actions yet. We didn't define this control yet. Therefore, it is in progress status, not completed. We can face different examples in various production types. But if we understand the logic, it is quite straightforward. This was the end of PFMERview lecture. Now we will continue with the case study. 20. Case Study - CAPA for Occurrence: Let's start the action plan with occurrence root cause. Here is the root cause that we did for lack of change management. So our actions should cover all these points. The ease way is always starting from the technical root cause, and it is the second line. And the action we will take according to technical root cause, will be corrective. So it is change management procedure for temporary tool adjustment were not clearly defined. So let's take an action for this. Change management procedure will be updated. Rollback process will be added and temporary tool adjustment will be detailed. This action fixes that root cause. This is corrective. We also identify the responsible deadline regarding the root cause column. It is just to show the link with the action, so we can just put a reference number of root cause rather than typing go. Now we will take an action against the systemic root cause which is just underneath. It has lacked awareness to anticipate the risk associated with temporary tool change. Here we should think about what can we do to improve this point? This will absolutely be preventive because it will prevent the technical root cause while fixing the systemic root cause. Let's identify conduct risk awareness training on how temporary tool change can create unexpected risk. While this action is addressed the system of root cause, we can also add one more to standardize this. Let's say, implement a simple risk assessment step before applying any temporary tool change. With this action, we will have standardized this robustly. Let's bring the other occurrence root cause. It training scope was not clearly defined. To fix this, we can define a corrective action like revise the training scope to ensure punch alignment verification and related failure risks are fully covered. It's pretty comprehensive action. We also define action Owner as Atak and the due date. As part of this action, we also organize a training for punch alignment to ensure all relevant teams have the necessary knowledge according. Now let's bring the systemic root cause to remember. There was no structured qualification or competency requirement for those preparing training content. Here the question is, how will we prevent this in the future? In a different project with a different team, we can make the same mistake because we didn't put some standardization yet. Preventive actions are exactly for that. If you identify some requirements to the system, we can prevent this to reoccur in the future. Let's write it. Define qualification and competency requirements for training developers. This action will provide a standardization because we will prevent incompetent people to prepare training content. Now, we completed the corrective and preventive action plan for the occurrence root causes. Now let's move on to the next lecture of study and create action plan for the non detection. 21. Case Study - CAPA for Non-Detection: We'll create an action plan for the non detection. Let's bring the root cause analysis that we have done in the case study. The verified cause was lack of calibration. Let's remember it. We had a control fixture for checking the whole of the semi products and the control was being done in that way. If the fixture passes, then the whole locations are okay. But the problem was that we identify in the FDA analysis, this fixture was not in place because it was sent to calibration. Here in the root cause analysis, we should take action for both sides because we have two separate root cause for the lack of calibration. Let's start with the first section. Here, our technical root cause was production startup approval process hasn't got the full equipment list. If you also read the previous case just above it, number three, it says fixture availability control is not included in production startup checks. Here, as a corrective action against this technical root cause, we should update that production startup approval process to include full equipment list and also we should put some checklist to ensure those equipments are checked before starting to product if you see the number five, which shows the systemic root cause, as there was no formal procedure to ensure all equipments are checked, we should develop some procedure or policy to address this. Let's speck to the action plan now. First action as already considered for the technical root cause, update production startup approval process, and create a checklist of all required equipments. This is a corrective action and it covers the technical root cause. Let's see the other actions that we have. Develop standard operating procedure. This action source the underlying reasons of the technical root cause or in the other word, it address directly the systemic root cause. Therefore, this is our preventive action. Here, we also put one more action as training for the operators and supervisors about the NIV procedure to ensure everyone understands the NIV procedure. This action is part of the action above, identified to support it, it is preventive as well. So we hit one more root cause section in the same analysis about lack of calibration. Therefore, we will also get some actions about this one. The first root cause, which is a technical, no contingency plan in place for alternative control options. So we should create contingency plan for alternative control options in production process. And regarding the systemic one, which is just underneath it, since there is no documented requirement to establish alternative inspection methods, we should put some documented requirement to standardize this application. Now let's see the action plan what we can put in place for those lag points. First action against the lack of contingency plan is create contingency plan and identify alternative control options such as CMM or visual inspection. Since the relevant cause was about the lack of control fixture, because Vasanto calibration and the production continued with that control fixture, this action will provide an alternative control when the control fixture is not available. So if you give an example, it will be like, if the relevant control fixture is not available, use three D CMM. This is the alternative equipment to make this check. Basically just an alternative control, which is also mandatory for the control plans. So this was the first corrective action. Now let's see the second one. It is provide training on when and how to use alternative inspection methods. This action is a training action, and it will provide a basic understanding to the controllers. So since it addressed the same root cause and support to just action above, it is a corrective action as well. Now let's put a preventive action to address the systemic root cause. It is ensure that contingency plans are included in the control plan. Control plans are the fundamental document of production processes. Those are the standard documents that includes all the process controls. This action states that we should identify an alternative control for each single control item in the control plan. This is fairly enough for the standardization. Now let's do this for the last root cause of non detection. It was about the assembly fixture deformation at the process. If you remember, how was it, as you see? This was the assembly fixture which was used to assembly of two sempducts as well as the whole control with the pins. But the problem that we verified in FDA, it was the middle pin of the fixture was missing. So when the defected part introduced that fixture, it was passing. That was the second cause of non detection. Now let's remember the root cause of it. The number four, which is a technical root cause, maintenance instruction doesn't have a control step to ensure all fixture components have been checked and assembled properly. We should update this maintenance instruction accordingly, and we can also create some checklist for the maintenance operators to ensure. And regarding the systemic root cause, which is the number five, maintenance procedure doesn't emphasize the necessity of enduring fixture functionality and condition after the maintenance. Here, we need to update this procedure and add this as a requirement. Now let's back to the action plan. The first action is updating maintenance instruction, which address the technical root cause. Therefore, it is a corrective action. And the second action to develop a standard checklist for maintenance operators to ensure all fixture components are correctly installed. Again, this action is also corrective as it still address the technical root cause as an additional action after the one above. And the last action is update standard maintenance procedure to highlight the necessity of fixture verification. This action is addressing the systemic root cause, therefore, it is preventive. 22. Case Study - PFMEA Review: FME review. Now you see a complete process FME of the case study part. It includes each process of the part. Let's bring the process from the case study. First process is stamping. It is where we produce the first semi product and the line items on PFMEA are here, blue highlighted ones. The second process is stamping again. This is the second sempduct that we had a quality claim for that part in the study because the second hole was misaligned. The third one is assembly, where we assemble these two parts. Their PFME section is also here. Now let's find our issued line where we have the failure mode of hole misalignment because that was the failure mode that we received from the customer. So this is our line. It is under second semi product, piercing Corporation and under assembly hole position requirement. The failure mode is hole position is out of tolerance, mislocated. This is exactly our focus point to update. Now let's go step by step. Regarding the potential effect, bolt cannot be assembled or fors. This is still valid because that was the issue at customer line and they couldn't assemble the part. Now let's see the potential causes of the problem. They are misaligned piercing tie or punch and tool were. We started updating from this point because we had a few root cause of that problem. The first cause is piercing punch wrong placement and the second is inadequate control of engineering change process. Regarding the current preventive controls, it was not exist, so that means no control. Therefore, occurrence of these points are quite high. We rate it as seven. According to the FMEA occurrence table. Detection controls are still same. We didn't change or add a new control and the reason of it, it wasn't unsuccessful because of the control type or frequency. It was unsuccessful because of lack of calibration and maintenance of this control fixture. All this is about the effectiveness, not the control itself. Therefore, we rate it as four, which is the same considering the above controls. Because the mechanical pin action which is written above already applied, so we will not consider that like a new action. We can directly write here in the nv line as part of the current action. The RPN is 224, it is quite high number, but now we will reduce this because we already planned corrective and preventive actions. These were specific training activities for the punch alignment and rollback process for the temporaries. In the new rating at the right side, as you see, severity and detection same, but we reduce their occurrence rating according to the occurrence table. Their new RPN is 96 and the second one is 64. So now we update the PFME according to factors that we identify in FDA 55 root cause analysis and the actions. We didn't bring all the root cause and actions here because, for example, the contingency plan action is just about the alternative controls, which should have already been defined on control plans. On the other hand, the control fixture maintenance checklist action, which was planned to ensure all control fixture components in place and properly working, just about the effectiveness of current gauge controls, which were already defined. So we didn't change anything about those. That was the end of the lecture. Now we will summarize corrective and preventive actions. 23. Summarize D6 Corrective & Preventive Actions (CAPA): Corrective and preventive actions, we basically first listed the root cause, both for occurrence and non detection, including the technical and systemic root cause. We identify corrective actions against the technical root cause and for the systemic root cause, which are the lack of standards, we identify preventive actions that provide standardization. Developing of action plan can be seen in one of the easiest section of eight process because once the root cause are identified, only thing that you should do is put some actions to fix this. That is basically correct. However, the key point is actions should cover all the lack points through the root cause analysis because root cause analysis can just state that lack of procedure. However, we should also identify action training after creating a procedure where it is necessary. Therefore, it is very important to develop this action plan with the relevant stakeholders to cover all relevant areas. Other point is actual date. We just identify the actions. However, once they are completed, we should record its actual date here in the same table. One other thing is action deadlines. Due days. The deadlines can vary depending on the action type, their difficulties because while some actions can be completed in a few days, others can take few weeks and sometimes even months. But the priority is always try to complete as soon as possible in line with the determined dates in the action plan, of course. In that point, we need to agree for the action plan and dates with both action owners and customers as well. Once we identify the action plan, we need to review and update PFMEA accordingly, which is another key point in the step. This was the end of corrective and preventive actions. Now let's make a quick quiz to reinforce our learning. 24. D7 Introduction to Action Validation & Effectiveness Check: The previous step, we had identified the corrective and preventive actions against the root cause. And in this step, we will validate those actions to ensure they have been completed and working effectively. Our first step will be the verification of the actions. We'll simply verify whether they are completed and working effectively. Second step will be the reproduction of defect, which is the other key point. We will try to reproduce the defect by intervening the process to ensure the defect can still be occurred or not after the actions implemented. Now let's dial in and understand how we do these checks. 25. Action Validation & Effectiveness Check: Our first step is verification of the actions. This is quite straightforward because we just go to the right place and check to ensure that the actions are in place and working effectively. We should observe the actual place and actions, we should get unnecessary details from the production to really be sure that everything is okay. This is a very key point because your limited time to check might not be enough to see some disruptions if there is. It is very useful to get feedback from the process supervisors or operators at the relevant process. After this check, if everything is okay about the actions, we can confirm. This verification should cover all the actions in the action plan. For example, if this is a standard document action, then you should check the updated content and sureties embedded than the relevant system. Or if it is a physical action in an operation process like here, then you should go to the real place and check. Second step is reproduction of defect. This is crucial because in this point, we try to reproduce the defect to ensure whether our actions address the root cause and solve the problem. We had already applied reproduction defect in FDA analysis, if you remember. But at that time, the aim was injuring the potential cause is really causing the problem itself or not. So we had simulated the failure mode by providing same conditions according to the relevant line item in FDA. But here we will do this to ensure that we cannot produce a defect anymore. After the actions, it is to confirm that our actions addressed properly. Because in this step, D seven, if we still able to reproduce defect by manipulating something in the process, then that means our actions are not effective. We should turn back and evaluate our actions in that case. Let's give some practical example to understand. For example, problem occurred due to changing of machine parameters, and our actions were about keeping stable of these machine parameters. So in this point, we should try to manipulate those machine parameters during the production running. So if we can change it, then that means our action is not effective. Defect is still able to so we should turn back to actions and evaluate again. Another example, let's say our problem is wrong placement of the part during assembly operation and we apply the Poka yoke to avoid this. In that point, we should try to put the part in wrong position again to see the result. If Pokoke is effective, it will not be possible to manipulate it, which is our expectation. We just try to simulate the conditions. Let's give another example. The problem is different reference of part welded in welding operation and our action is adapting sensor control for part identification. For the reproduction of defect, we can try to put different part there again to ensure that sensor is detecting it or not. If it detects the wrong part, then we cannot produce a defect, which is the expected result in this step. Basically, our aim here is to ensure that we cannot produce a defect in no way after the implementation of the actions. For the AT, we can document this like defect reproduction study has been done and failure mode cannot be produced after the implemented detections. We just tick the second validation action. However, there might be some cases where defect reproduction trial is not possible. For instance, if the issue is due to a design failure and the design has been changed or if it involves a die problem that requires maintenance, it may not be feasible to reproduce a defect. In such scenarios, once the die is modified once the modification is done and no longer produced a failure, diverting the die to simulate the failure mode is not possible and not make sense. Thus, this reproduction of defect approach generally applies to conditions where we can manipulate the production process. Another important point in these works, we should provide the necessary conditions to avoid any mistake because reproduction of defect is a risky operation due to working out of the standard. Therefore, it must always be under control by relevant supervisors or managers from the relevant operation team. For example, all the parts that produce in this period must be isolated because they can be defect. Another important point is health and safety, of course. It must be managed always by the teams who are authorized for the relevant process. During these shop fuller controls, we can also conduct process audits which is more comprehensive and apply some failure mode analysis studies such as diverse FMEA. Kind of applications are more structured, but can vary depending on the organization you work in. So in case you do, you can document these actions like process audit was conducted, no NC was detected. Actions have been verified, which is for the action verification. And for the second, verse FMA study was conducted, no issue was found. Failure mode cannot be reproduced after the implemented actions, which is for the defect reproduction. But at minimum, we should verify the actions in place and ensure that defect cannot be produced anymore. And sometimes we might have to monitor the action effectiveness for a time period in some cases. For example, process has been monitored for four weeks, no issue was detected, and this kind of monitoring actions might be considered if the defect is very rare and singular. So you might want to check the process in a time period in such cases. But this is not always possible in the automotive industry, especially, because obviously, you need to complete all study in a very limited time and submit to your AD to the internal or external customer. This was the end of the lecture. Now we'll continue from the case study. 26. Case Study - Action Validation & Effectiveness Check: Regarding the case study for action validation, the first thing that we will do is verifying the actions. It is to ensure these have been applied and working properly. Let's bring an action from the previous step. The first action is updating change management procedure and adding rollback process with temporary tool adjustment. Here, we directly check the procedure to ensure these requirements are added. We checked it and verified that. All necessary items, document updated properly and embedded into the quality management system, which is important to provide document control and then cascade it properly. Perfect. First action done, we verified this. Let's bring another one from the occurrence action plan again. This is conduct risk awareness training on how temporary tool chains can create unexpected risk. We can demonstrate this by getting the training attendance report. System always needs a demonstration. We need to have a proper way to demonstration according to our actions. We check the training content and verify that. It covers the points of how temporary tool chains can create risk. We got the training attendance report, so we verified this one as well. Let's bring one more from the non detection action plan. It just develop a standard checklist for maintenance operators to ensure all fixture components, including pins are correctly installed. This is a document action. However, it should also be checked in the maintenance area, in the shop photo area because this is a form and it need to be used by the operators. We need to ensure their access to this and their understanding on how to use. We check this and verifying that form created, all necessary components added to check, and operators have access to form in the chat floor. Now, in case any maintenance, they will check the fixture conditions to ensure all components are in place and correctly installed. Perfect. We verified this action as well. We are doing this verification properly to all the actions both for corrective and preventive. Occurrence and non detection. After that, we document this verification on 80 step D seven. Actions have been checked and verified in Gamba. Gamba is a Japanese terminology, that means actual place. Instead of shop flow or the actual place, gamba can be used. Now we need to ensure that the actual defect cannot be produced anymore. We will check the shop floor. Let's remember the defect occurrence first. It was whole misalignment. This was the press and this black part is the one arm of bontinge. Operation is done and pierce three holes, three aligned holes. This was the normal condition. Now let's look at how did it occur, how the failure occur. As you see, the second hole is not align in the second visual. It is not in the middle of the part like we did in the left visual. The reason was second punch in the tool was incorrectly placed during an engineering changing and it was not reverted after the trial. Afterward, we investigated the root cause and took necessary actions against this. But for the reproduction of defect, this is not a failure mode that we can reproduce by manipulating the process. Basically, this case is not feasible to directly reproduce the defect. However, through a simulation during the tool maintenance, we verified that the defect cannot be reproduced after implementing the new actions because we put specific checks and tool rollback process in place. Consequently, we validate this step as defect cannot be reproduced after the implemented actions. 27. Summarize D7 Action Validation & Effectiveness Check: D seven action validation and effectiveness check step, we verify the actions by checking them in action place. We verify whether the action is in place, is working properly. If there's a standard document action, the relevant teams are aware of it. Another check was the defect reproduction. While it was not always possible to make this trial directly in proper cases, we should try to reproduce the defect to ensure our actions are effectively working and avoid defect reoccurrence. Other point was about the checking method. We can conduct process audits or reverse FMA study to confirm the production process, the actions, and also the current and potential failure modes. This was the end of this section. Now time for a quick quiz. 28. D8 Introduction to Lessons Learned & Closure: So far, we addressed the root cause with corrective and preventive actions. We check the action effectiveness and now we have come to the end. In the step, we will close the 80, but we need some additional things to do before this. First, we will capture lessons learned. We should document the lessons that we learned from this eight study. This will basically summary of problem, root cause, and implemented action. The importance of it, this will be considered in the NIV projects. For example, they will design the process or product by considering these lessons learned. Lessons learned are one of the significant know how of companies. Therefore, they are the fundamental of this step. Afterwards, we'll make 80 closure audit to ensure all steps completed properly, and then we'll complete this 80 by celebrating the team. 29. Lessons Learned: Lessons learned. As the name suggests, it's a lesson that we have learned through the process. This is a deployment and a promotion tool to capture the key outcomes of a problem solving study. This is quite straightforward. It is simply a before and after statement of the action. Let's make an example to understand this better. So now you see an example of a lessons learned card. At the top left side, we have registered information like Lessons door number, lessons door type, whether it is coming from occurrence or non detection, 18 number, which is to understand from which 80 and also part or product number. And next to it, we summarize the problem. And after that, we bring the root cause analysis in FFiveFrame and finally, we put the visual of solutions in before and after format and indicate the implementation action just as below. This is a simple format of a lessons learn card. Let's review what does it tell? This is a non detection sss learned as indicated at the register table. Problem was sharp bur on the screwing hole of one part. It was detected by the customer at the assembly operation and this bur caused a minor injury to the operator. Let's review the root cause analysis through to five pis. It Burr was missed during the inspection. Why? Because operator didn't detect a part during that visual control and why? Because it was too small to detect as visually. The next question is why was checked then as visually if it is difficult to detect. The answer is because improper control method selection. The selected control method was not proper to detect this failure mode. Clear. Then we can ask one more why because the risk was not predicted in PFMEA, process failure mode and effects analysis. So what we have done against this route because as seen on the picture before we had a visual control, but now we adapted a camera control to detect the birth. As a summary, camera control adapted and PFME control plan also updated for this failure mode. This lessons learned card or LL card in an acronym tells us, visual control is not sufficient to detect small burs on the part. You should adapt a camera control rather than a visual and the risk needs to be considered in PFMEA. This is what we understand from these lessons learned card. Now just imagine you are starting a NIV project in the production in the same production plant, and you will produce a similar metal part like that. Instead of putting a visual control for this failure mode, you could probably consider to adapt a camera control because you learn to lesson. Visual control is not effective for the small burs. Lessons learned are one of the main input of the NIV project. It is important to capture this to avoid future problems. 30. 8D Closure Audit: ATO ort is the final step in our AT process. In this point, we will ensure that. Did we complete each steps properly? This is a self assessment of all the process. We will check each point and if we have some gap, we will go back and correct it. So we have five questions for each discipline and simply we just assess the results. If we meet the criteria, then we select ok. On the other hand, some of the questions might not be applicable for the 80. In those cases, we just put NA, not applicable and we might face some failures for some of the questions if we don't meet the criteria. For those, we just put noto for those, we need to fix this by using a simple action plan in the same form. And once there is no not okay, the result will look as accepted, which means 80 can be closed. Now let's go through the questions. For the D one, we basically check, did we use five WTH? Do we have visuals for the problem statement? Do we have problem definition both for customer and supplier? This is on applicable if the issue comes from the customer. So if we do this AT study internally for a problem regardless of customer or any other stakeholder, then this question will be just NA. And the last one, did we notify teams in the company about the problem, which is a simple mail share or can be a QRQmiting, maybe AI, if the company work in N principles or any other communication method according to the organization that you work. D two simply checks three main points about the similar parts and processes. If we miss the evaluation of the other similar process or other version of the parts or customer plans, we should turn back to D two and update it. D three, is the initial analysis questions. Which are about the shop for basic checks or gamba walk in the other we ensure we had initial analysis for non detection, which was required for a robust containment action plan and basic condition checks in the process, which are like machine parameters are okay, operator trainings are okay. Is there any anomaly in the process, et cetera? So some basic initial checks. D for containment action. First, we check a defective or suspected parts, table prepared, which is for dirty batch analysis, and then sorting activities to ensure no bad part sent to the customer. Afterwards, containment action. Here we ensure that containment action plan developed and completed. Next is the five. Here we check heavy done fishbone analysis, fault analysis, and five fives root cause analysis for both occurrence and non detection. As the last question about the systemic root cause here, have we formed the systemic root cause through the five fives analysis if applicable. To document the systemic root cause is not a mandatory task in AT studies as long as the preventive actions prevent the root cause happening again. However, it is extremely important to unfold underlying systemic issue about the lack of standardization. Next section is the six corrective and preventive action questions. Here we basically check that our corrective actions address the root cause and our preventive actions prevent that root cause. As a last important point here is PFMEA process failure mode and effects analysis. We should check to ensure it was reviewed and updated. The seven is action validation step, and here we ensure heavy completed corrective and preventive actions and check their effectiveness to they work effectively. There are two questions about the capability and measurement system analysis in this section. These are not always applicable. Regarding the capability, the index is CP and CPK. If the problem is not singular, let's say, and if there are some variation issue about the part dimensions, for example, we should consider to have some capability analysis to verify that process is capable. Or let's say if we did some improvement or changing on the part geometry or in the process directly like changing of the machine, for example, we should ensure whether process is still capable. But other than, these are not required always. Next is MSA in this section, which is necessary for the verification of measurement method by constrt the variables like operator and part as well. Again, this is not always required. For example, if you adopt a new camera control method to the process like a camera control or a gauge control, for example, we will need to verify those with MSA. But other than if it is not relevant, then it is not required. Last question in D seven is reproduction of the fact. We ensure that this study has been done and defect cannot be reproduced anymore after the actions. And last step is D eight. Here, we should ensure that lessons learned captured properly. So once all is done and if there is no Nok status, the result will be acceptable and we will be ready to celebrate the team to close the AD. 31. Team Celebrate & Recognition: Team celebrate and recognition. At this stage, you can just relax because all steps have been completed. Here, as the 80 leader, we ten people in the team, recognize their efforts and contributions and close the in the step, people tend to just close the 80 and go away. But a recognition, which might be a small thank you via mail or a recommendation comment on your company database is really important to keep people motivated and let them know that you aware of their effort. One other thing at this point, if you receive this problem notification from the customer or from an internal audit, let's say, then we should submit this 80 to get their approval to close. This might be a customer for the external quality problems or maybe a sponsor for the internal quality problems. Once you think to people who work in the eight study and get approval from the sponsor or customer, then you can close the 80. 32. Case Study - Lessons Learned & Closure: Case study of last step D eight, we will start with the lessons learned. We had a few root cause in D five, that means we should prepare these LL cards for all of them. But in this case study, we will just make one to keep it simple. Let's have a look. We identify the basic information at the top left area like part number, eight number, lessons learned type, and number, and bring the problem definition here with a summary of five WTH and bring the root cause from D five step and identify the visual statement of the actions as before and after. At before we already had change management procedure, but there was no rollback process and no risk assessment which are identified in the root cause analysis as well. Now in the corrective and preventive actions, we improve these processes, so updated this procedure. We clearly show this statement in the after with the updated procedure. Now we have a lesson learned card for our eight. In the AT audit, we simply review the questions and answer them according to our 80. For the initial steps, D one, D two, D three and D four, we identified problem with 5w2h, both customer and supplier. We used visuals and published the problem. We had done similar parts and process analysis very detailed and we conduct initial analysis at the shop floor. Finally, we had developed content action plan with a defective part table, so all is okay. Regarding the D five, we made a very detailed root cause analysis, but we didn't make reproduction of defect for FDA causes because it was not possible to revert back to tool to produce bad part. So all this section is okay, but the reproduction of defect in FDA is NA. And regarding D six, we developed corrective and preventive action plan and implemented properly. We also updated PFMEA. So this section is also okay. In these seven, we verified corrective and preventive actions, and we also tried to reproduce a defect in the process. So all the actions were in place and they were working well. Regarding the capability and MSA, these were not applicable for our case. So they are NA. And the last question, lessons learned are also captured. So that's it. There is no gap and the result is accepted. Now, all this big study done, we completed the full 80 and received a customer approval for IT. So we than our team, or stakeholders and close the 80. So if we populate the steps respectively in the eight, first we confirm that if any further standard updated including FMEA control plan, maybe product specifications or other main procedures. This should have already been done in the six corrective and preventive actions. Here we just identify these updated standards as reviewed and updated. Lessons learned created a T audit completed without a gap, and team recognized, and customer approval has also been received. This shows that 80 is closed now. 33. Summarize D8 Lessons Learned & Closure: D eight, we started with lessons learned where we captured the problem definition, root cause and implemented actions with before and after visuals. Lessons learned were important for know how, which is key to success for following projects in the future. And after that, we conducted 80 closure audit to ensure each step in 80 problem solving study is okay. And next, we completed the 80 by celebrating and recognizing the and of course, getting customer or sponsor approval where it is required. This approval was also important because if you had a quality problem raised by an external or internal customer, you cannot just close problem solving study without getting their confirmation. If it is fully internal and inside the department only, then quality manager approval is enough or the one who is assigned by the organization. Once all is done, we populated the final step respectively, the first FMEA control plan and other specification update confirmation, and then lessons learned, at the audit, and team recognition, and the customer's approval. 34. Congratulations!: You have just completed the course on master at problem solving root cause analysis, D five to the eight, so a big congratulations. That means you now know how to perform effective root cause analysis, define corrective and preventive actions, validate their effectiveness, update PFMEA, and close problem solving studies with lessons learned. And more importantly, you now have the confidence to apply these tools in real world situations. If you found this course helpful, I would really appreciate if you could leave a quick review, even just a few words. And also, if you haven't get yet, don't forget to check out my previous course, which covers the first part of the AT problem solving study, D one and D four. Thank you again. I wish you a continued success in your professional journey wherever you are in the world.