Steps in the Improve phase
– [Voiceover] We’ve reached the fourth phase in Six Sigma project, the improve phase. What is the purpose of the improve phase? Well, at a high level, the purpose of the improve phase is to address the proven key Xs, and come up with solutions to improve the Y. What does this mean? Throughout this course, we have been using the example of a pizza chain with customer complaints about crust. Let’s say the proven causes, or key Xs are oven temperature, baking time, and tossing technique. So, in the example, the question to be answered is, what is the best combination of solutions for temperature, baking time, and tossing technique? The sequence of steps in the improve phase are as follows: Generate potential solutions to address proven key Xs, evaluate solution alternatives, select the right set of solutions and implement. However, before implementing the solutions, you must apply these tools and techniques. Process maps of improved process. FMEA, or failure modes and effects analysis, used for identifying and mitigating any potential risk of failure in the new process. Mistake proofing, to error proof any possibility of errors occurring, and pilot testing prior to full implementation. Various tools, such as design of experiments, DOE, and process simulation may be used to determine optimal settings. For example, in our pizza class problem, DOE can be used to determine the optimal settings for oven temperature, baking time, and tossing technique, as well as methods and procedures to reduce any variation from those settings. The resulting optimal X settings may be 426 degrees Fahrenheit, 11.2 minutes in the oven, and tossing the pizza up three times clockwise. Methods and procedures are developed to ensure those levels are set correctly, and any variation is reduced. With those solutions for the proven Xs, there will be a reduction in pizza crust problems, resulting in fewer complaints, fewer lost sales, and fewer refunds to customers, and let’s not forget the organizational side. On a changed management note, it is critical that key stakeholders are involved and engaged during the improve phase. Key stakeholders include process owners, process operators, managers, and others who are impacted by the problem. Their input and buy in are critical. Ideally, these folks would have been engaged throughout the entire project, either as project team members, or as part-time subject matter experts. So, again, the steps in the improve phase are: Generate potential solutions to address proven key Xs, evaluate solution alternatives, select the right set of solutions and implement. We’ll be covering these steps in more detail in this chapter. Understanding the purpose and steps involved in this phase will help you develop effective solutions for the proven key Xs to improve the Y for your project.
Generate, evaluate, and select solutions
– What is that old saying about insanity? Doing the same thing over and over again and expecting different results? Such insanity is exactly what you don’t want to do when you develop solutions for your project. Let’s discuss how you should generate, evaluate and select solutions. To generate ideas and potential solutions use brainstorming and other creative thinking techniques. One such technique, taken from Edward de Bono, is Six Thinking Hats. This technique calls for members of the team to wear different colored hats, where each color represents a thinking role such as optimism, devil’s advocate, creativity, and so on. Basically, the person wearing that hat takes on that personality. Another technique is Anti-Brainstorming where you brainstorm how to make things worse. The project team shouldn’t be the only ones coming up with the creative solutions. You want to get the help of subject matter experts, process owners, and process operators. Get all the help you need. Once you generate a list of possible solutions, the project team evaluates the list to come up with a shorter list. Techniques used for short listing include Multi-voting and 2X2 matrices such as effort impact or cost benefit quadrants. So, those solution ideas requiring little effort and low costs that have a high impact and high benefits should definitely take the short list. The short listed ideas are flushed out and detailed solutions developed. Where possible, solutions are evaluated scientifically using such techniques as DOE, or design of experiments. In the pizza example, a DOE can be done to evaluate the best combination of settings for oven temperature, baking time and cost and technique. As you move from evaluating to selecting potential solutions, use techniques such as a Criteria selection matrix, the Pugh matrix and Cost-benefit analysis. The Criteria selection matrix evaluates and scores each solution alternative against agreed criteria weighted by its relative importance. The set of solution alternatives with the highest score is selected. In the Pugh matrix, selection criteria is used to compare alternative solutions against a baseline solution. Is it better, worse, or the same as the baseline solution? The alternative with the most data and the fewest worth is selected. Cost benefit analysis also factors in to the selection process. Obviously, the benefits have to outweigh the costs. Once you have generated, evaluated, and selected possible solutions, your work is still not done. The recommended solutions must be approved by the champion and other key senior managers before implementation. Following this process and obtaining their buy in will prevent you and your team from doing the same thing over and over again.
Reduce the risk of failure through FMEA
– Think about the process of applying for scholarships to college. What can possibly go wrong? Wouldn’t it be great if you could anticipate and mitigate any risk of failure? Well, the good news is that there is such a tool to help you do that. It is called FMEA or Failure Modes and Effects Analysis. FMEA was first used in 1960s for the Apollo Space Program. Can you imagine how many things can possibly go wrong with a rocket launch? A lot. Since then it has been used in many industries, and it is even a requirement in the automotive industry. It is also widely used in Six Sigma projects. There are two types of FMEA, the Design FMEA and the Process FMEA. The Design FMEA is for reducing the risk of failure associated with a product or service design, such as when you are designing a new hotel, or the next smartphone. The Process FMEA is for reducing the risk of potential failures in a process. I will focus on the Process FMEA because it is more relevant to Six Sigma demake projects. Let me explain how the Process FMEA works with a series of questions. For each step of the process what can possibly go wrong? In what ways can it fail? These are called potential failure modes. And for each potential failure mode, what is the effect? And, how severe is the potential effect, using a one to 10 scale where 10 is the worst. That is called the severity score. What are the potential causes for each failure mode? How likely is occurrence of these potential causes on a one to 10 scale where 10 is the most likely. That is called the occurrence score. What process controls are currently in place to detect the cause or the failure mode? What is the likelihood of detection on a one to 10 scale where 10 is the least likely to detect? That is called the detection score. When we multiply severity times occurrence times detection, we get a composite score called the Risk Priority Number, or RPN. The RPNs can be used to prioritize the failure modes. The highest scoring failure modes are the first targets for improvement. During the improve phase of a Six Sigma project, FMEA can be used to prioritize and mitigate the risk of failure. Process steps can be improved, and process controls put in place to reduce the RPN. It can be tempting to think that your improved process will not go wrong, well the truth is it will. But what you can do is to reduce that risk by using FMEA.
Mistake proofing
– When you open a microwave door while it is heating your food, the microwave shuts off immediately. Thank goodness it does. That’s mistake proofing. Would it be useful to have mistake proofing in your processes, products and services? Let’s discuss the basic principles or mistake proofing, and how they can be applied during the improve phase of a Six Sigma project. Mistake proofing, or error proofing as it is sometimes called, is best when it prevents errors from occurring, and if that’s not possible, the next best thing is to facilitate the work so that errors are minimized. Lastly, if errors do take place, then detection should be obvious and immediate, or automated. There are basically three levels of mistake proofing. Here they are in order of preference. Prevention, Facilitation, Detection. The most preferred form of mistake proofing is the prevention of errors. A perfect example of this is the way ATMs or cash machines used to work. Remember the sequence? First, you insert ATM or debit card. Enter your pin code. Request cash, say $40. What came out first? Cash, then receipt. What came out last? Your card. As a result, after getting the cash, customers often forgot their card, and many cards were left behind, benefiting the next customer. At that time, I was traveling a lot, and I came across some machines with a slightly different sequence. The first thing that came out from the machine was the ATM card, then the receipt. And the last thing that came out was the cash. So there is no way I could have forgotten and left the card behind. So, a simple re-sequence of process steps mistake proofed any possibility of leaving the card behind. No additional cost involved. Just a simple sequence of steps. Of course, the technology and machine has changed since then, so that today you just slide your card. No insertion required, card problem solved. Another form of mistake proofing by prevention. If prevention is not possible, then mistake proofing by facilitation is utilized. It is facilitation of work to minimize errors. An example of this principle is the cash register at fast food restaurants. The keys or buttons show a picture of a hamburger or large fries. And the cashier simply presses the picture without worrying about typing in the correct price. This minimizes errors, but it does not prevent them from pressing the wrong buttons by mistake. If mistake proofing by prevention or facilitation is not possible, then utilize mistake proofing by detection, where detection of errors is immediate either by the obvious or automated. An example of this is automatic spell check in documents using word processing programs, where wiggly lines are drawn immediately after a word is spelled incorrectly. So, integrate mistake proofing into your improve process. Where possible, apply mistake proofing by prevention first. If that’s not possible, then apply mistake proofing by facilitation and detection. This way, you’ll prevent errors, minimize errors, or at the very least mitigate it’s effects, immediately.
Pilot test and implement
– How often do you dip your toes into the water to test its temperature before diving into the pool? That is similar to doing a pilot test before a full-scale implementation of improvements. Let’s discuss the importance of pilot testing and planning, prior to implementation. First, what is a pilot? A pilot is trial run, a dry run, or a small-scale test to make sure the improvements will work as planned. It is an opportunity for the Six Sigma project team to understand how the improvement solutions will work on a very limited or small-scale basis, before a full-scale implementation. Pilots enable any kinks or problems to be uncovered and fixed before the project team rolls out improvements plant-wide, division-wide, or company-wide, across all locations. The importance of doing a pilot cannot be emphasized enough. For example, when improvements are made to a machine or a process, it is best to pilot it first in one production line, or in one location, before implementing it across all 100 locations company wide. It may come as a surprise with many folks, but a pilot is actually a data collection exercise. Why? Because data is collected during a pilot so that answers can be provided. Answers to critical questions such as, how well will the improvement solutions work? Will it show enough improvement? Will it work without any problems? If no, what are those problems, so that they can be fixed and the solutions revised. What can be done to ensure a smooth, full-scale implementation later? Those are just a few of the many useful questions that will be answered by a pilot. In addition to pilot testing, planning prior to implementation is important. Planning the implementation with regards to communication to ensure buy-in from key stakeholders, including what is the right message to communicate, by whom and to whom. Procedures that need to be revised. Responsibilities and authority that may need to be changed. Training, who to train, when to train, how to train. Timing of the implementation, how should they be coordinated or staggered. Time and effort required for the implementation. Budgeting for the type of resources and level of resources required. In short, an implementation plan is required. Planning is critical, because if you fail to plan, then you’re planning to fail.