Zen and the Art of Manufacturing – Episode 1 ft. JT Badiani

Transcript

Bryan Sapot (00:05):

Hi, this is Bryan with Zen and the Art of Manufacturing podcast and I have with me today JT Badiani, who is the president of Focused Improvement. JT, if you don’t mind kind of saying hi and give us a little bit of your background and we’ll jump into it.

JT Badiani (00:18):

Great, thank you for having me today, Bryan. As you mentioned, JT Badiani. I’m a chemical engineer with a background in Lean Six Sigma, significant background. I worked as a President of Focus Improvement, but also operated a number of different companies. And I’ve been working in Lean and Six Sigma for about 30 years now across multiple sectors from food to automotive to… durable goods to aerospace to pulp and paper and pharmaceuticals. So variety of industries and great experiences.

Bryan Sapot (00:52):

Yeah. what we had, thanks for that. And what we had talked about discussing today was something on everybody’s mind at the moment, which is tariffs, right? And that kind of how they’re changing and they’re really outside of manufacturers control. And so the thought was talking about some strategies based on, your background, lean principles, maybe AI on how companies can reduce costs and maybe become more efficient to deal with whether. know, tariffs that are going to be charged on their goods or tariffs they have to pay on incoming goods. So with that kind of give me your thoughts around what you’re seeing or how you’re advising companies these days.

JT Badiani (01:29):

So yes, tariffs are at the top of mind of many of the companies I’m working with, predominantly all of them. So I have one company that I know of that 99 % of their goods go into the US. And they’re really stepping back and trying to figure out how they manage to the tariff process. On the other end of the spectrum, we have another company that is set up so that about 20 % of their goods go to the states. So the conversations are unique in that spectrum. The philosophies that I’ve been using is, hey, know, whether there’s tariffs or no tariffs, we do need to take a look at your processes, take a look at your data and figure out where we need to make improvements. So go back to the grounding of lean principles, you know, get a value stream map, understand where your bottlenecks are. understand what your constraints are, make sure that you’re managing those constraints to the optimal and then try to fix those and reduce the waste along the way. So that’s been the core tenet of my messaging. A lot of companies know that. They’ve been trying to do that, but they forgot. Because, hey, look, in Canada, you can pick up the exchange, which helps out with the cost of the product. You know, if you’re doing well and you’re growing and your margins are there, yeah, a little bit of waste is not a problem. But now things are getting tighter, right? The owner of that company that has about almost 100 % of their product going to the States, they’re freaking out right now. They’re really scared about what’s going to happen and how they shift. So I met with them a little while ago and started pushing really hard on, let’s look at data. Let’s capture the data. understand what’s happening. They’re a food manufacturer, so they’re hoping that they can implement some technology, start capturing the data that’s coming off their lines. They have a lot already, but it’s not in a format that you can use it. So we’re going to step by step go through, analyze where the weak points are, and then start to drive some change in our organization. So that’s… Sorry, go ahead.

Bryan Sapot (03:40):

Is it? So is that like the best place to start is kind of get a baseline set of data on your lines or equipment and then make changes from there?

JT Badiani (03:50):

So from my perspective, yes, like I tend to be very data driven. A lot of the owners have started to develop a strategy. So, you know, like some of them are saying, Hey, we got to shift away from, you know, reliance on the U S if you’re looking at it from a Canadian perspective. If you’re looking at it from, you know, from a U S perspective, I would drive and say, yeah, let’s look at data. Where are the weak points? Where are we missing the opportunities? Where can we take headcount out and start using whatever technology they have to drive the improvement? Some companies are still behind the curve and their head is in the sand and they’re struggling to figure out where they go and how do they grow either business outside of North America and what do they need to do? So it depends on where they are. So consultants answer, my apologies, but it depends on where they’re going. and where they are in their journey. The one thing that is very, very unique in the conversations I’m having is all of them are bringing up AI. And you mentioned that earlier, right? So I’m to kind of deviate towards that for a short while. So I attended a session last week on Thursday, and we had a couple of speakers come in and talk about AI and the benefits of using AI in industry. And one of the key things that came out was, know, AI will be able to help us in a number of different ways around looking at data, you looking at ways of capturing the right information, translating it from, you know, verbal to text and vice versa, right, or into videos. Putting it into action is another story. Like, you know, if you’ve got Gmail or if you’ve got, you know, Microsoft Outlook, yeah. Co-pilot and Gemini will do the things to summarize emails for you. That’s great, right? That’s fantastic. It saves us time to looking at and reading a bunch of emails. But how can we take that to the shop floor? If you’re an operator running a piece of equipment, that’s not going to do anything for you. So one of the things that we’re starting to work with is, okay, let’s take a look at our back office processes. figure out what we need to do so that we can go to that next level. It’s like, Bryan, you and I have probably gone through this, right? We’ve gone through and said, hey, you know, can we put PLCs, you know, out on a shop floor on pieces of equipment to automate this data capture, right? Then can we automate the process so that it moves on its own? Can we put a robot in? We’ve gone through those changes over time, right? And AI is just another tool that goes above all of that. And we got to figure out how to use it. And that’s, We’re starting to come in and try to help companies figure that out.

Bryan Sapot (06:42):

So what do you, how are they using it in the back office? Like I think you were mentioning that there’s some specific examples where you have a customer that’s doing it and seeing some good success.

JT Badiani (06:51):

So they just implemented Dynamics 365, Microsoft Dynamics 365, and they have Microsoft CRM. So they’ve got a CRM, they’ve got D365, and their processes are fairly robust. And we were looking at one of the processes yesterday and said, okay, how can we use AI to help manage an order coming in through to ordering the parts? We started working through the process map. We then took their ISO certified, took their procedure, put it into an AI and said, here’s our process map. Here’s how it’s documented. Help us optimize and reduce the cost or the time to make this process, execute this process. And the AI came back and after about five or six iterations asking different types of questions. we got a full blown plan as to what to do. And it was amazing, know, with Power, Power, sorry, with Automate, with AI and use of Power BI and some Python code, there were some patches we needed to do and AI spit those out. We can now take a sales order, put it from, sorry, a quote, convert it into a sales order, convert it to an engineering. build a material once engineering finishes their stuff, upload the bomb, then, know, purchasing all they have to do is hit a button. And that whole process can be automated pretty much. The creativity part is the engineering and the sales. Everything else will be flow through, right? And A, that’s going to reduce errors. B, that’ll help us, you know, drive improvements in. and cost by taking that person that was doing, or two people, doing certain roles, transitioning them into a different part of the business now.

Bryan Sapot (08:43):

So to do that, just to kind of dig into the details, because sometimes it gives people inspiration on how they can use this themselves. So you guys put together this value stream map and some documentation around it, it into one of the bots, like ChatGPT or Copilot. Which one?

JT Badiani (08:58):

Yeah, yeah. We actually did chat GPT and it was the enterprise version that we used. Okay. And as I said, the prompts that we, there was multiple prompts that we went through, right? So went through chat GPT. We then did the same exercise in co-pilot, right? So we took the same prompts, reworded a little bit, cause we learned along the way, put them into co-pilot and tried to figure out what co-pilot would come back with. There were some, I would probably say 98 % of it was the same, but there were some nuances along the way. Co-pilot’s learning model is a little bit older. It’s a little bit dated. So it didn’t have as much of the newer information, we even asked it, what would it cost to implement this change? And so Co-pilot came back with a higher price. It was 2X what ChatGPT came back.

Bryan Sapot (09:51):

Interesting. So has this been rolled out yet? Are they still working on

JT Badiani (09:55):

So in the next couple of weeks, we’re going to flush out the plan. We have two strategies that we’re working on right now. One is, we go and get a company to come in? There’s a bunch of companies in the Toronto area that can help us implement. So we’re going to probably talk to them. And or do we hire an internal resource? We thought about hiring a co-op student to come and do that for us. And that’s a discussion we’re going to have in a couple of weeks once the plan has been flushed out. Do we go on hire? Keep that resource in house, that knowledge base stays with us, and then go on go and train another person internally to manage that.

Bryan Sapot (10:32):

much time we think that’s going to free up once it’s in.

JT Badiani (10:34):

Um, so, uh, really good question. Cause that was the one thing I was really worried about was what’s the hour of walking on this, right? Now we, you know, we, we might see, um, headcount reduction depending on where we land. Um, but the time that it’s going to save us is probably going to be, um, about per order between eight to 12 hours, depending on the size of their mate to order. engineered product company. So it’s kid-mounted products, know, pumps, vessels, things like that, valves. So, you know, it depends on how big their bomb is. So it’s going to vary a little bit, but it looks like it’s going to be about 8 to 12 hours.

Bryan Sapot (11:12):

That’s interesting. Yeah. I mean, depending on the number of orders per week, like that’s a lot, right? That’s a lot of time saved.

JT Badiani (11:14):

Right. Yes. So when we did the math, it’s probably equivalent to about 10 to 20 orders a month that we get. So if we take the lower number, let’s say 10 times 8, that’s 80 hours times 12, that’s about 1,000 hours. That’ll come back to you. And just imagine if you take that resource, those hours, and deploy it to something else. There’s an opportunity to do other things in the business now.

Bryan Sapot (11:47):

Yeah. Well, and then use the same model to map other value streams and see what else we can automate. Right.

JT Badiani (11:54):

So it’s funny, so let me tell you the other use case that we looked at, because that’s exactly what we did. Like after we did the first one, we started looking at the accounts payable process. And, very similar exercise. We haven’t flushed it out to the very end, but, you know, we’ve gotten to the point where we said, okay, here’s our process map. We learned from the purchasing process, right? And said, okay, how are we going to do this now? So we put in the steps that we do it. And again, Power Automate and AI were able to do a bunch of things for us. Now, the one thing we are making sure we don’t do is the issuing of checks or e-transfers. We want have a wall there that says, we’ve got to have a person look at this because we don’t want it to go wild. So we’ve put a hard stop there.

Bryan Sapot (12:49):

Have you tried to extend any of this stuff out to the factory floor in any of your clients? has anybody done some little pilots on that?

JT Badiani (12:58):

So one of my clients from last summer, we helped them implement technology on the shop floor, very similar to what Nulogy does. So they just tested looking at, they have a pick and place robot, just a very simple XY, picks it and drops it. So we just started looking at taking that data, putting it to AI and seeing how we can optimize. anything in that cell. Okay. So, we have the data from the PLC and the equipment. It’s a welding process. So, part goes in, gets welded, picks it up, drops it into a box, and then the robot kind of does a spin circle. So, we just got that data. We’re going to see what we can do with it. I do have a use case from another friend of mine who does a lot of work in… an AI for one of the larger firms. And he took data from a mining company and put it through their AI model. So they used a proprietary AI model that they’ve developed. The basis is still chat GPT, but they’ve added another layer to it. And they took a look at the ore output. So if we can get, or if they can get, you 1.1 % more ore coming out of a mine through their process. That’s a significant impact and benefit to the mining company. And so they’ve optimized their process now so they can get out several points more or points several tenths of a point more now. And it’s just by looking at data optimizing it and using AI to come back and say here are the set points in the ore process. I don’t understand mining at all. I’m not a mining expert, but He was telling me that’s a massive improvement in their throughput.

Bryan Sapot (14:55):

Yeah, it’s interesting. So we have a customer, they’re a big auto parts customer that make parts for Subaru and they had kind of a line balancing problem in terms of like, how should we schedule changeovers and stuff like that, optimize the number of people. And he, the general manager of this plant used to be an IT guy. So he’s like a data person. And so he extracted the data out of the system, our system, you know, formatted it correctly. built a prompt, put it into Microsoft Copilot and said, here’s my constraints, here’s the lines, here’s the standard cycle times, all that kind of stuff. Here’s my demand, load level it, right? And it did it. And it it blew his mind. It actually gave me some ideas from a product perspective, but it’s something that they would have to do. You know, it was an Excel problem before you would use solver and that kind of stuff to try to figure it out. And it would take hours. This took them about 30 minutes. So.

JT Badiani (15:42):

Yeah. Au

Bryan Sapot (15:47):

It’s pretty cool. They don’t do a ton of changeover. So once this load balancing model is set, they’re going to continue to use it that way. But it’s really interesting. Whatever you can dream up, these things can do it. One of the other things that I found in our work internally is you can, and maybe this is obvious to everybody, but you can use the systems to help you improve your props. So you can do a prop and then say, OK, make this better for you this is trying to the output I’m trying to get and then it’ll blow out all the requirements for you and then you can adjust it and feed it back into the system and get better results than you know like a one sentence thing.

JT Badiani (16:26):

So it’s funny, we did that and it told, was, we had it set to sarcastic mode, okay, for a while. So really interesting, the chat GPT output we got, because it was really sarcastic, you know? And then so we played around with it a bit and then we turned it back to normal mode, right? Getting back to your level loading problem, one of the… use cases that we’re going to try the next couple weeks. So this factory, another client I’m working with, they have a weld shop, electrical shop, assembly, paint shop, several processes that they’re trying to schedule. so what we’re doing right now is doing exactly what you just did, is understanding what their constraints are, what their demand is, and then we’re going to put that all into a spreadsheet. and say, schedule this for us. What work orders should we be doing? I worked with that fabrication manager and we said, are the things that we need to weld. And we put all the weld bins down and said, estimate how many hours it should take. So we started working at that level to see if Chad GPT could tell us how much effort it will take to do this type of weld. So we put a couple of their projects in and it came back and the hours were pretty good. We let those projects go to the floor. We knew what hours that chat GPT said. We compared it to what the lead hand said it would take to do that work and what sales did. So we have three data points. Now we’re just waiting for the actuals to come in and we’re going to adjust the model and say, Hey, you were this much higher or this much lower. Now use this actual data and run the schedule for us and tell us. what it’s going to go through. And we’re going to go through the whole process that way. So get actuals, pump it back in, and then see if it comes back with the schedule. The goal, the ultimate goal is how can we schedule the plant so that we get the right action happening at the right station? Because what I know is when I walk through is I’ll see a guy standing there and I’ll ask him, hey, how come you’re not doing anything? Like, yeah, you should be sweeping or doing something, right? And he’s like, yeah, I’m waiting for that part to come out of that, know, weld cell to come over here. And I’m like, okay. So how can we get it so that your downtime, your effort is maximized and you get value out of his day?

Bryan Sapot (18:58):

Which is always the challenge in those multi-step processes. There’s some constraint that keeps everybody upstream or downstream standing around. So let me ask you a stupid question. Because of the industry that you’re in, you’re helping coach people and instill these best practices around lean and problem solving and continuous improvement. I almost feel stupid asking this question, but it’s like, ChatGPT has all of this knowledge too, right? So is it, how do the two work together, right? So like if I’m a manufacturer and I can get all kinds of different knowledge out of ChatGPT on how to maybe I should optimize things and do different things that we’re talking about now, where do you think consulting falls in that world to augment what I can get out of these systems? Does that make sense?

JT Badiani (19:43):

Yes, so there’s actually two thoughts or two questions with that, okay? Because I’ve been thinking a lot about this and you’re absolutely right, right? All the books that I have behind me are useless because ChatGPT has it, right? And so there’s two things there. One is taking a concept and understanding it and then having a creative solution. That’s where consulting has to end up, okay? ChatGPT can do that front end just like that. Like I said, 30 minutes, 10 minutes, 30 seconds, it’s gonna get a response. And it’s learning all the time, so it’s getting faster and better. But translating that into an action on the shop floor is what we both bring because we’ve been around the block and we understand people, we understand how to motivate, how to coach, how to lead, how to translate that thought into an activity. The second part of this is, and I’ll be more transparent on sort of where my head is at is how do you charge for that? Okay, so the thought leadership that we used to have is gone. Hey, know, JT knows this, Bryan knows this, and we can charge for that knowledge, right? So now you turn around and go, okay, it’s a different type of thinking that we have to apply, and how do you charge for that? it adds value to the customer, right? And so I’m adjusting my costing models or our costing models so that we can say, look, if you want to use Chatt GPT to the front end, great. Now let’s talk about how to implement this. And it goes back to that age old concept of, the consultant knows where to hit the machine with the hammer. It goes very quickly, right? As opposed to, you you go and hit it multiple times, right? That’s what you want to avoid.

Bryan Sapot (21:27):

right. Yes. Right. Right. Right. Yeah. I paid 10 grand for somebody to show up and hit a machine with a hammer. It took two seconds to fix. You didn’t pay for the two seconds you paid for the 30 years worth of experience the person has to know where to hit it with the hammer. Yeah, exactly. Yeah. I see it as a, almost like I haven’t thought deeply about it in terms of, you know, the consulting industry, but like you have to… And we talked about this in all of our stories before. You have to have good information to put into it, to ask good prompts, to get good questions, to get good answers out of it. And a lot of that takes some expertise. You also have to know, you also have to be able to validate the response, right? Like I’m sure you’ve played around with it enough to know that when you ask it a question about something that you happen to be an expert in, and then you look at the output, you’re kind of like, it’s not quite right.

JT Badiani (22:19):

Yeah.

Bryan Sapot (22:20):

but you ask it something else, you know, where you’re not an expert and you’re like, great, it’s like gospel, right? It works. It’s, we’re going to take, take you out to be these works, a word for it. I am, it’s like an augmentation thing. It’s like you said in the beginning of the podcast, it’s like they’re tools to help us do work better. So it’s like, it’ll make you more efficient. The client will probably be more prepared when they engage with you, right? Cause they’ll have a better understanding of maybe these concepts and how they can apply them. And maybe they’re stuck, right? with that, you know, how it all meets the floor, right? Yeah.

JT Badiani (22:53):

100%. Yeah. So, there’s a young leader that I’ve been working with the last little while, extremely bright guy, his shipper receiver left. And so, you know, he emailed me and asked me, hey, you know, what are some options? And I said, see if ChatGPT can do it for you. Right? Like, if you can avoid hiring somebody that’s going to do this paperwork. Can you not find a way to upload it to chat GPG to print it out for you? And it’s done. Right. And so, so I have to follow up with him as to see, you know, what, it can do and what it can’t do. Um, but, it’s coming back to, you know, having, having the, the idea of saying, let’s try something in a different way. Right. And that’s what, that’s what we bring to the table. You know, don’t hit this thing with the screwdriver and you gotta use a hammer. Let’s talk about what type of hammer you want to use. Alright, let’s figure that out.

Bryan Sapot (23:56):

So what, mean, you know, to come back to the top on talking about, you know, tariffs and stuff outside of our control. And this is a really good time to, think invest in efficiency and things like that. I even, I listened to this economics podcast and even with all the tariff craziness that’s going on, they’re still projecting in most industries for manufacturing to continue to increase in margins, to increase that kind of stuff through 25 and 26. And they’re even telling folks to do the same thing, like gear up and get prepared for this, because it’s going to keep coming. The economy should still be doing well unless there’s some crazy black swan event. anyway, what are some of those? And I always like to talk about the simple things, because I think it’s always hard to get started. What are the other perfect places to start? Where do you look? Where’s the best place to look for trying to find efficiencies? We talked about data. But what else?

JT Badiani (24:53):

Yeah, great question. So I was at a factory today, right? This morning. And they make epoxies and chemical products, right? For multiple industries from construction through to battery production and circuit board. They cover the circuit board with epoxy and also do a bunch of castings. That was the exact question that they asked is where should we start, right? And so when we were walking in the factory floor, there was one area where there was probably about a dozen to two dozen workers, know, kind of huddled around doing a bunch of work. The other areas were like, you know, three, four, five, six people and the process was doing its thing. And so, you know, I’m standing there, I’m talking with the CEO and I said, you know, I see something, hub of activity here. what’s going on, right? And he started explaining his process and you could easily tell where the most people were was where the constraint was. And so that’s where I kind of asked him a bunch of questions around, know, cycle time, people, two shift, one shift, and how do you manage a constraint? Do you have a schedule? So what I would recommend is figure out where your bottlenecks are, where your constraints are, and then, you know, work your your lean, whatever tools you want to use around implementing improvements and throughput, focus there. Okay, start with your constraint. Because if you can fix your bottleneck or plan your bottleneck and control it, then you can say, I gotta go two shifts, so I gotta do this. And like I said, if you are expecting growth, three, four, five, six percent, wherever it is, make sure that you can manage that bottleneck that much better, right? And then once you get that happening, the bottleneck is going to shift. We know it’s going to go up or downstream depending on what’s going to happen. Make sure that you have the tools to recognize where it’s gone and in the same thing, start working around that. One of the things that I noticed was he didn’t have or they didn’t have work orders with all of their finished goods. So they have a schedule, but I don’t know how it’s communicated. So the visual factory was the other thing that I talked about was how does the operator know what to go to next? Is it a clipboard? If it is, I don’t see it anywhere. So if you’re manage a constraint, start using some basic tools, get that evolution starting to happen. The other thing, simple things like 5S, we’re keeping up, cropping up along the way. So start with the constraint, go from there. Okay. The other thing I asked them was, and that would be one of the early things I would do, is do they have a value stream map? Do they really understand how their process works and what their changeover times were, what’s their attack time, all those things that are important to understand when you’re looking at it from a grant scale and then start picking those projects that come from that BSM.

Bryan Sapot (28:01):

The thing that I’ve always seen with the VSMs is it seems like such a daunting thing. Especially if you’re like, if you’re a really high volume manufacturer and you run the same stuff down like five different lines and you have like two changeovers a week, it’s easy, right? Like you can do it in like 30 minutes. But if you’re like that company you were describing that’s engineered to order custom pallets, like it could look like a spaghetti diagram based on. like, how do you start with that one? So it doesn’t feel so overwhelming to begin with.

JT Badiani (28:32):

What I’ve typically done is a lot of those factories, would go down to the, call it the area level. Start with electrical, for example, or fabrication. Let’s do a VSM for that. So you do it in subgroups. Then you go up to the higher level and then you mesh it with sales data, SNOP data, whatever you’ve got.

Bryan Sapot (28:49):

Okay.

JT Badiani (28:59):

And then you bring the higher level together. Because one of the things we don’t want to do is improve a process that might be a dying product line. You don’t want to improve that. Let’s go back and say, what’s selling? What’s working? And then that comes from leadership and then S &OP. And then you come down and say, OK, this is where our bottlenecks are. Let’s start with those areas. So I tend to do it at the subgroup level and then go to the larger. company level or the group level.

Bryan Sapot (29:29):

So is it just to kind of parrot that back to you? it’s kind of, okay, look at it, look at the sales and operation level and like what’s really working. So we’re selling a lot of this, right? And then we’re going to lean into this and probably sell more of this, whatever product it is. And then look down that level into the department to go, okay, where’s that bottleneck or like where are the problems that we’re having? And let’s focus on those particular areas, build out the VSMs for that. then, yeah.

JT Badiani (29:56):

Yeah. All right.

Bryan Sapot (29:58):

And maybe you even just improve that before you move on.

JT Badiani (30:00):

Exactly. And then you would iterate, right? So let’s say your bottleneck shifts from fab to assembly because now you’re putting more throughput to assembly. Okay, now let’s figure out how to fix that or understand that it’s going to impact you. Get the assembly team working on some improvements because if you’ve got the departmental level VSMs, then you can step back and go, well, that’s a bottleneck, that’s a bottleneck, that’s a bottleneck. Let’s work on this one first. prioritize this one, this one’s number two team, get ready, because it’s coming your way, right? And then as that leadership in that area starts fixing that process, as the fabrication improves, for example, then Fab can pick up and so on, right? And then it comes down to going back to business and the strategy, which should come from your annual planning process and say, here are the product lines we’re going to work on or here are the products we’re going to work on. let’s make sure that those pieces of equipment or those lines are set up to accept that higher throughput.

Bryan Sapot (31:03):

Yeah. Okay. Makes sense. And so then just to keep drilling down like into the details. Okay. So I have my area of VSM, right? I know that let’s just go with fabrication, right? We know it’s a problem. How do I, what am I doing next? guess like the VSM itself may expose some inefficiencies in the process that we could just fix like, right. But then below that, I mean, are we like, doing time studies, like out on the floor, watching what’s going on, try to figure out like what the details are of where these issues like take me through that a little bit.

JT Badiani (31:34):

Yeah, so exactly that. is it a time study? it, know, go through the Tim Woods, right? So, you know, is it a transportation issue? Is it inventory? And is it motion? Is it overproduction? So you start going through the Tim Woods, right? And go, okay, I’m looking at this cell or this area. What are the issues, right? You know, is it the labor isn’t trained or you don’t have proper procedures? know, standard work isn’t… It could be as simple as they don’t know how much to produce by the hour. It could be just as simple as what’s the throughput required by the hour, right? Or how many linear inches of welding do I need to do per day per welder? So it’s just setting those standards and expectations and saying, hey, here’s what we need to do. But you’re absolutely right. I would start using spaghetti diagrams. you know, Tim Woods, would take a look at, you know, their time studies, if they have any, do they have any standards, what’s their standard work look like, if they don’t, you know, then you get a list of improvements. And then what I typically do is I’ll sit back and I’ll talk to the guys on the shop floor. They know what’s happening the best. So you talk to them and you talk to the supervisors and then you start working with them and saying, Here’s all the things that we think we need to improve. This is from your team, this is from you, and this is from us, our thoughts. Let’s understand where the biggest impact would be. And if it comes out to be time studies, then let’s get that prioritized, right? Or sometimes they say, I want to invest in a new forklift truck, or I want to invest in a table that can do this, that, and the other thing. Okay, great. That’s a $50,000 investment.

Bryan Sapot (33:08):

That’s what we do.

JT Badiani (33:22):

Is that going to have the same value as something else with lower or no investment? Then you start doing the trade-offs and you put that called short, medium, long-term plan together. And then you walk away to management and say, I need money for this. And here’s the short-term things we can do to get the wins. One of the things I always coach young managers or young leaders on is you have to earn your way to the purchase of new capital equipment. You can’t expect to go to leadership and say, want an X dollar device, whatever it is, you got to earn your way to it. So you have to show that you’re going to, you’ve taken waste out, you’ve created this much extra capacity dollars wise, and then I’m going to spend it over here. So as soon as they give you that check to go buy the equipment, get it in and make sure it runs well, and you earn your way to the next opportunity.

Bryan Sapot (34:18):

Makes sense. What’s Tim Woods? I haven’t heard that before.

JT Badiani (34:21):

Tim Woods. So it’s a lean acronym. It’s all the all the wastes in lean, right? So transportation waste, inventory waste, motion waste, overproduction, know, defects, skills, I’m missing one in there. Overproduction, I think I got them all, but those are the Tim Woods acronym that comes through.

Bryan Sapot (34:46):

That’s good one.

JT Badiani (34:47):

No? Okay. I gotta have a sit down. I’ll train you next time.

Bryan Sapot (34:51):

I’m not a lean guy, right? I’m a software person who’s lived in manufacturing forever. So just absorb this stuff over time and try to help build tools to enable it. What have you seen from, this might be the last thing we talk about, but from a tracking perspective, let’s say you do the map, you figure out the areas that you really want to focus on. How are people running those projects and tracking them?

JT Badiani (35:18):

So larger organizations, set up a PMO, a project management office, PMO, and usually you put one or two or three people in there. And their role is to do multiple things. First is make sure that all the tools and templates are available. Usually it’s a SharePoint site that they’ll set up and put everything there. They also track the dollars that are saved. I don’t necessarily mean that they go on the floor and say you saved this much that has to come from the ops group But they collate all that data and then make sure that it’s valid. So they do a mini audit Okay, they work with the finance or accounting group to make sure that the dollars are real and then the third thing they do is they set up I Asked them to set up the certification process for you know for lean Six Sigma. So if you’re a yellow belt, green belt, black belt, they facilitate that process and they monitor the training or set up the training piece. So that PMO office for larger organizations helps drive the momentum. They also help set the strategy for the business, where to focus next, things like that. there’s some other value that they add to create that drive for continuous improvement. If it’s a smaller organization, 20 people, 50 people, 100 people, typically those are the clients that I work with, the PMO office doesn’t work. So I give them a couple of options. I can take care of that with you. I’ve got the team that can do that, that’ll help you manage that. And that becomes sort of the momentum to drive change in their organization. Or we get one person, typically it’s like the quality manager, they… spend a bit of their time and they do that. They work with the ops team and that’s typically where it lands. And they’ll work with finance or accounting to do that. But you slim it down so that it’s not, I don’t want to make it a bureaucracy. That doesn’t add value. But for a smaller organization is to say, did that project save you money? How much did it save you? Let’s validate it. Move on. We need one or two people to manage that.

Bryan Sapot (37:27):

And I guess that would be on larger projects. Like if we’re going to make a big process change or we’re installing automation or something like that. What, what do you see on more of like the day to day, week to week, month to month, little things, like how are people managing and tracking that stuff?

JT Badiani (37:27):

Thank So I’m a firm believer of visual, whether it be now digital is easy, right? Because TVs aren’t that expensive, our monitors aren’t that expensive, and capturing data is quite easy to do with technology. So what I like to do is take a look at it from a productivity or efficiency perspective or OEE, whichever metric you want to use, and say, hey, let’s target where you were before the improvement started and where you are today. Right? And just simple things, throughput or, you know, as it mentioned OEE, you has that improved? Right? And start with that. So if you’ve got a baseline, you go up from there. There’s a company I work with, they do window screens. Okay. And they measure units produced per labor hour, UPLH, which is pretty cool. Right? Right.

Bryan Sapot (38:39):

Yeah. Yeah.

JT Badiani (38:41):

That’s a really good way of looking at it, right? And so before we started doing all the value stream mapping and such, we baseline their UPLH. And over time, we want to see that jump up. And it goes up and down because you’ve got seasonality in that business and orders dry up and if housing industry slows down. And so there’s a lot of dynamics there, but that’s their constant. They know what their UPLH is to be productive and efficient. and the contribution margin for the company needs to be at a certain level and they’ll tell you right away, okay, where it needs to be at. So all I would say is measure what you think is important, typically it’s throughput, OEE, productivity, efficiency, but just make sure you have a standard measure for everybody across that department or that plant.

Bryan Sapot (39:32):

and then drive and measure, yeah, improvement based on that metric. And don’t change the definition of it.

JT Badiani (39:35):

Yeah, and you do that plan, check, plan, do check act and you start going through that nickel over and over again.

Bryan Sapot (39:42):

Yeah, it’s interesting. Like we find that. Like manufacturing companies, know how to fix stuff, right? The challenge that we run into with Nulogy Smart Factory is like, okay, we give you the data, give you the Pareto chart. They look at it in the meeting and says, here’s the biggest down times, right? Across these machines. And then there’s like this disconnect between like, okay, now what? Right? Or like, how do we look at this systematically every day, every week to see if we’re driving improvement there? Which is that it’s… PCDA a little bit, but then it’s also, do I interpret and drill down into the information? Which is actually probably, it’s more of that. Like how do I draw good conclusions that I can believe in and then go take action on? Cause I can do the actions, right? I got the people, I know how to do it. Like we’ve been running a manufacturing company for a long time, but it’s like, how do I use the data to help figure out where I should, where I should go? And folks struggle with that a little bit.

JT Badiani (40:34):

So you’re absolutely right. It’s translating that data that’s coming in and then saying, here’s the two or three things that I need to change in order to get the output. And that’s understanding your fishbone. So if you’ve done a fishbone exercise to say, hey, I want to increase throughput. Here’s all the things that can go wrong. When it’s this type of situation, we have to do this.

Bryan Sapot (40:49):

Okay.

JT Badiani (41:02):

And then if they also have a failure modes and effects analysis, they can also use that. So I tend to do productivity, it’s equation of all these inputs. Let’s take those inputs, figure out what can be problematic. So what are the root causes to prevent you hitting your throughput numbers? And then let’s get a control plan and then understand that control plan, develop it with the leadership and it’s typically the supervisors. Because they’re