Episode_10-Herb-Ready

Herb Robbins, the founder of Lean 2020, discusses the importance of control plans in continuous improvement. He highlights that control plans tie together various tools and provide guidelines for holding teams accountable. Without a control plan, processes can become disorganized and less effective. Herb explains that control plans should include key characteristics, both inputs and outputs, and should cover everything from critical to minor items. He emphasizes the need for data collection and automation to improve accuracy and efficiency. Welcome to Lean into Excellence, a Workstream Consulting Podcast. I'm Liz Crescenti. And I'm Marco Bonilla. And we will be your hosts as we embark on our continuous improvement journey. Welcome back to another episode of Lean into Excellence. I'm Liz Crescenti. And I'm Marco Bonilla. And today we have Mr. Herb Robbins, who's the founder of Lean 2020. And we will be diving into the world of control plans. Welcome, Herb. It's great to have you here. Tell the audience a little bit about yourself and your journey to continuous improvement. Sure. Thanks, Liz and Marco. I appreciate the time to take out of your day to welcome me here and talk about control plans. So a little bit about my background. It's relatively diverse. I graduated from a marine engineering school in Baltimore after spending a year at sea in three different merchant ships. I visited over ten countries by the time I was 22, received a commission in the Navy, and I had my third assistant engineer's license, steam and diesel, with unlimited horsepower. I also have a BS in mathematics from Charter Oak College in Connecticut. When I graduated, there wasn't really a lot of options on the water. So I was fortunate enough to land an entry-level engineering position with one of the most highest tech facilities in the world relative to building printed circuit boards. So really fortunate. It's actually one of the flagships of today, TTM. And again, they're the largest manufacturers of electronics in the world. I spent 24 years in a variety of roles, process engineering, R&D, quality, manufacturing, operations, and mostly at the single division. And then I took over the operational excellence program in around 2003, and that's when I had responsibilities across the U.S., Europe, and Asia. Around that time, I earned my Master Black Belt with Tyco International and SBTI. And I left Tyco and TTM, and I started my consulting company 17 years ago. And haven't looked back. It's just been an incredible journey for me. I've dealt with clients in close to 50 countries, and I do mostly trainings as well as consulting. And I'm going to give a little plug out to my website, lean2020.com, where I offer, you know, white, yellow, green belt training. I can do live and remote trainings, and been fortunate enough to work, again, with a lot of different companies. So some of those companies are in healthcare, finance, other sectors. Plug out to NASDAQ, the stock exchange. I've done work with them in the UK and in France and in Indonesia. I've also done work with a lot of military and aerospace. So that's my strength is military, aerospace, electronics. I also perform some NADCAP and MediCred audits in my spare time when I don't have anything else to do. And that also gets me around the world quite a bit. On a personal level, I have two adult children, both launched successfully with two grandsons. And I've helped start two businesses with my son. And using many of the tools of Lean and Six Sigma, I have to say that sometimes kicking and screaming, I've exposed both of my kids to these tools and been quite successful. So I'm proud of them. Additionally, I like to play a little bit of pickleball, living in Florida, the nicer weather during a good part of the year, and do some biking. And I competed in my first CAP Ironman with my son this past fall. Awesome. I mean, Herb, you stay busy. So thank you for being here today. Yeah, I'm either traveling the world or training for Ironman. So yeah, I'm not sure how we locked them down to begin with, right? So this is amazing. So Herb, seeing so many different companies, what tools do you see as important drivers of process control and improvement programs? Terrific question, Liz. So during the earlier podcast that I've been listening to with our fellow consultants, I've heard process mapping, I've heard measurement system analysis, qualitative versus quantitative tools, and even process capability. All of those are on my list. And there's certainly relationships that tie all of these tools together. However, my favorite is going to be the control plan. And that's why I wanted to talk about that today. When I go into a company, and one of the first questions I'm asking when I'm performing either a process audit or some type of review, I'm always asking, do you have a control plan? It's a hill that I live and die on. And I really want to stress the importance of that today. Herb, I have to ask you, why a control plan and not another tool? So for me, a control plan ties a number of these tools together. It's like one-stop shopping. We identify process inputs, outputs, acceptability criteria, responsibility, reaction plans. There's so much more that a control plan can do. This is a document that we create and can be used within all levels of an organization, providing references and being able to hold a team accountable. Those that have been to some of my trainings know that I talk about accountability as a four-letter word, something that we don't like to use four-letter words. Accountability is not easy. It can be difficult, even awkward to hold someone accountable. However, I believe the root cause of holding people accountable is management's responsibility and their lack of defining roles and responsibilities well. So what a control plan will do is a control plan will help provide those guidelines that will allow management, whether it's upper management or even middle management, the opportunity to identify, hold people accountable to those roles and responsibilities, and make sure that we're monitoring a process properly. Hey, Herb, in your experience, do you see improvement essentially fall apart if there's no control plan in place? Yes, absolutely. And I will say it will vary quite a bit, but one of the things that I have encountered in many organizations are things that are documented in multiple places. I've literally seen up to five different locations for, say, a temperature, pressure, some process controls, and it would simplify things to just have it in one place. Any time it's in two or three different places, the chances of it being the same drop dramatically. So having a good control plan will, again, establish those controls in one place that can be easily referenced by other documentation, and the success of a program or project is increased immensely. Herb, can you provide some examples? Sure. So depending on whether your process is transactional or operational, there's certainly things in the process that we're going to be able to pull out. So product key characteristics would be something that we measure directly on a product or thing going through the process. So it could be a part thickness, a bearing radius. In health care, it could be a patient's blood pressure. In manufacturing, it could be a hole diameter, a length or a width, anything that we're measuring or assessing on a product. A process key characteristic might include an area's temperature or humidity, could be a pressure, conveyor speed, could be a machine number. If you have three or four x-ray machines, we want to identify which x-ray machine, what potential x-ray technician was part of that process. We can also measure things like backlog or yield, cycle time. So there's a lot of different things within the process that we can do. And an aspect of the key characteristic is also whether it's an input or an output to the process. Inputs to the process are very important. Those are what we call leading indicators, something like the temperature of the process or a part dimension that we're doing prior to the process. Afterwards, or our outputs, which we call lagging indicators, they need more things that we've already experienced a problem. And they're going to give us indicators as to what went wrong in our process, but still important that we're able to identify them so that we can stop our process, go backwards, and try to figure out what went wrong. Herb, it sounds like control plans can have lots of items. How do you know what to do? What I like to include in a control plan basically is everything. And if I've got a process that is, say, really important to get something out to a customer, just a document, I may not go into the weeds too much, but generally speaking, I like to control everything from the most important to the least important item. We get it on a control plan, and then later on, someone can identify something as maybe more significant or less significant than what I've originally identified it as. So I'll give you some examples, because what we don't want to do is we don't want to control everything to the highest level. We've all heard when everything's important, nothing is important. So it is important for us to clarify what I'll call critical to major to minor key characteristics. So a key characteristic that would be considered minor is something that we're not going to collect any data on. I consider it a process indicator. It might be a setting on a machine like a conveyor speed. It may be data that we talked about already about collecting a machine number. So it's just things that I'm verifying. I'm not going to bury those. It's, say, a set point on a machine. There may be some variability in it. But I'm just verifying that it's between a window area and I'm moving forward. And it's really important because these levels of key characteristics make up the majority of items in a process. And again, most of these things, we're just verifying, making sure that they're being performed properly. The next level would be considered a major key characteristic. These have strong effects on our process and product, and therefore, we need to not only verify them, but we want to be recording these things. And any time we can record things automatically, that's really important. So we can take the human element out of that measurement system. So if we can, again, collect data through some sort of automatic means, we're saving people time writing, maybe miswriting information. We've all transposed numbers incorrectly in the past. So we want to eliminate that as much as possible. We can also, after we collect data, something that Greg Massiana in an earlier podcast talked about was process capabilities. So if we collect data for a major key characteristic, and we certainly know what those requirements are, we can actually determine capabilities. So we can assess capability on those major key characteristics. So on top of majors, we have what I like to refer to as critical key characteristics. These are the most important characteristics. There's probably just a handful in most organizations. And again, we consider this almost like a pyramid where we have a lot of minors, fewer majors, and then a handful of critical key characteristics. These are on the top of the food chain. These absolutely affect the product and process quality. So in this particular case, we're certainly going to collect some data. We're going to track these statistically using our control charts and other tools. I will tell you that we will have done a measurement system analysis on these as well. And it's important that we look at them based on their data types, whether they're variables or attribute data. And again, we're going to do the appropriate measurement system analysis. So, Herb, you just started talking about this. Can you touch on the two types of data? Sure. Good question. And probably a good time as well to clarify this. So the two types of data that we have are variables and attribute. Variables is the type of data that we measure. So based on the resolution, the tool can provide almost an unlimited variety of measurements. So pressure, temperature, length, these things that could potentially have decimal places out to, again, multiple decimal places and, again, have just a large amount of possibilities. In the health care world, it would be a pulse rate or a blood pressure. So just a couple of those are just a couple examples of variables data. Attribute data would be data that we assess, and those would be things that we put into categories like pass or fail, good or bad. We could have different types of defects, different issues that happen during a review of a particular document. So those, again, categorical data is attribute. For a variable's data, we're going to talk about a gauge R&R. Can multiple people in a process measure and obtain the same results? That's our repeatability. I'm sorry, that's our reproducibility. Well, repeatability is within an operator. So I want to see is can a single operator repeat their measurement over and over again? In the world of attributes, we can show a number of different inspectors different types of defects, and we want to see the same thing. Can individuals repeat their inspections, and can multiple inspectors reproduce the same results? So those are some key things. A few podcasts ago, we had a gentleman, Kurt Peterson, talk about measurement system analysis. I'd encourage anyone here to go back and listen to that. These are huge, you know, again, and just another huge tool that plays a big factor in a control plan, especially when dealing with some of the critical and major key characteristics. Yeah, Herb, and just to loop back to Kurt, if you don't trust your measurement systems, your control plans are not going to work, right? I mean, it's so simple. No, that's absolutely a great thing. You know, we just assume because a gauge is calibrated that it's going to give us reliable measurements, and that's not the case. Calibration has everything to do with how accurate the unit is, but it doesn't tell us about the precision and the variability from one operator to another in different environments. So, yeah, really good observation. And we all know that systems drift, right, whether because of usage, whether because of time, just environmental causes, right? So, again, whether or not you have a good control plan is really dependent on those measurement systems and being looked at constantly. Right. And, you know, a common question that's often asked is, how often do you go back and review a control plan, much like a measurement system analysis? And we've got a lot of things that affect our day-to-day. So, I guess a good answer would be, you know, we're looking at our capabilities periodically, and if we have any drift in our capabilities, if we're looking at control charts or, say, a change in that capability, that's probably a good time to go back to our control plan and just make sure that things are in line and that we've, you know, captured everything to the right degree. Yeah, and just taking that a little bit further, if you're looking at – if you're trying to control continuous data versus pass-fail, I'm sure that 9 out of 10, you'd prefer to have continuous data to control, right? Because pass-fail doesn't really talk about any drift in the data. Well, again, you probably can see some drift in the data. If you're looking at attribute data using a P or U chart, you can see some shifting in the data. But to your point about variables data being much more powerful, you know, absolutely. Attribute data will tell us how good and bad something is, but variables data will give us the degree of how good or the degree of how bad that data is. Absolutely, Herb. Great. Herb, a control plan seems to cover a lot. Is there anything else you feel should be included? So, yeah, a control plan does cover a lot. And again, as I've mentioned before, one of the things that we've got to be really careful is documenting things in multiple places. So, control plans will be a good place to have the roles and responsibilities. So, we've identified what the key characteristic is, what the specification is. Some of the additional things might be who's responsible for collecting the data, what's the frequency of that data, you know, where is that data stored. And we can also record some initial capabilities such as CPKs, some MSA results to be able to refer back to later on in case that, again, the process drifts. Remembering that we're not necessarily running control charts on all of our CPKs. I'm sorry, all of our key characteristics, just the critical ones. And there may be some minor ones that we want to look at periodically. And it's nice to be able to know what those initial capabilities are. So, if a process is not capable, using a control plan, we can look to verify, you know, potentially what were the requirements, were those objectives met? And if they weren't met, are there some additional enhancements that we need to make? There might be a minor key characteristic that if a process is reviewed, we can say, hey, we're collecting some data on this. And potentially, you know, if this is a minor, we shouldn't really be collecting data. We can then determine whether that data should be discontinued as far as being recorded. Or we can say, you know what, this data is real important and let's continue it. But if we're going to continue to collect data, then we're going to upgrade that key characteristic to a major category. And ultimately, we can do the same thing. If we're collecting data and it doesn't seem all that important, we can discontinue that and downgrade a major to a minor and save people some time and effort. We don't like to collect data unless we're going to utilize that data. And it's not necessarily always for process control. Sometimes it's a regulatory requirement. There are people that use process control plans for their waste treatment. Maybe it's some financial process control plan. And it's something that we need to collect data for, you know, tax reasons or other legal concerns. So there's a lot of reasons to collect data, even though it's not necessarily directly affecting our process. Could be just an output. So, again, some really cool things to look for there. Herb, can you tell us what the best way or at least a mechanism of capturing the data for a control plan is? Yeah, something I haven't covered yet and and I actually hadn't even thought about covering it during this podcast. So that's a great question. One I've seen the most and the one I use the most, it's just simple, an Excel sheet. And that Excel has got obviously different columns related to each of the different elements of a control plan. What area is it from? What process step within that area? What's the key characteristic that I'm looking at? Is it a minor, major, critical? Is it a product or process? Is it an input or output? What are the requirements? So all those things that I covered are going to be part of that Excel spreadsheet. And it's important that as we're collecting that data, we may actually have a printout of that that we can use. But we're going to record that and that becomes the baseline and additional information such as that initial capability or MSA information will be in part of that. So having it as part of the software is really key. And again, it's something that you're going to be able to share across an organization, especially if you've got multiple locations that are performing the same operations, being able to share best practices and using a control plan for that is really important. Herb, are there any key takeaways that you want our audience to walk away with today? Sure, great question. So just wrapping up, I want to say that control plans can be used for any operation from manufacturing to transactional. I've used them recently with organizations to help document an ISO 9100 management review, clearly defining KCs to be reviewed, who's measuring them, what's the frequency, what are those goals and objectives, how are they measured? Those are all, you know, again, key elements of a good control plan. So just wrapping up, I'm happy to share examples of control plans. People can reach out to me at Herb Robbins at lean2020.com. And I'll send out an example using a simple Excel spreadsheet that I commonly use. And sometimes those even have links to procedures or quality records to help get people to other areas within an intranet, within an organization. So happy to share and hope there's some good takeaways from this. Yes, absolutely. Thank you so much for coming on the podcast today. And as always, you can find Marco and I at WorkstreamConsulting.com and Info at WorkstreamConsulting.com. And new episodes are released every other Wednesday. Thanks so much, Herb. Really appreciate it. Thank you, Herb. Great conversation today. Thank you. Very helpful.