Jay: AI, right, to me, Augmented Intelligence is the basic stuff. Forget about artificial intelligence, first should be Augmented Intelligence, right? How do you help the people to do things they could not do? 

Voiceover: You're listening to Augmented Ops, where manufacturing meets innovation. We highlight the transformative ideas and technologies shaping the front lines of operations.

Helping you stay ahead of the curve in the rapidly evolving world of industrial tech. Here's your host Natan Linder, CEO and co founder of Tulip. The Frontline Operations 

Platform. 

Natan: This week on Augmented Ops, we're chatting with Professor Jay Lee from the University of Maryland, where he serves as Director of the Industrial Artificial Intelligence Center.

While a lot of people love to talk about AI, Professor Lee has spent much of his career actually on the shop floor implementing AI. These tools and putting them to use in scenarios like building cars and many, many more. His career spans industry, government, academia, and it gives him a unique perspective on how we should be approaching this technology and its application.

Today, we'll be discussing with him how industrial AI is reshaping manufacturing operations, global supply chains, and how our education system needs to adapt to train the next generation of AI driven manufacturing practitioners. Welcome to Augmented Ops. Today we have with us Professor Jay Lee from the University of Maryland.

Jay and I have known each other for quite some time now, a few years. We've met at the Governor's meetings of the World Economic Forum, where Tulip was a tech pioneer. And Jay was like our teacher. Welcome to the show, Jay. Both, we are 

Jay: learning from each other. Yeah, 

Natan: we were just like preparing for this and I was like, wow, it's like very tough to summarize your very long and impressive career.

You've been in academia, teaching, you know, industrial operations. You have served in government bodies, you know, across NSF and of course, World Economic Forum. You build centers and nonprofits, but you've also been in industry. in companies like Foxconn and started their industrial artificial intelligence center.

How would you summarize your career? 

Jay: I think everyone's career all depends on destiny and also their purpose, right? Two different things. Some people want to do something, some people have a different destiny. The people you met, the people you knew might change our destiny sometimes. But I wanted to have a balanced career from day one.

So when I was in industry many years ago in the real world manufacturing, the automotive industry and later on in the service industry, you realize you want to do something that can Can make a big deal, but how do you make a big deal? So you need to understand from top down. Yeah. So I had an opportunity to work for, you know, National Science Foundation.

So you can oversee a lot of stuff going on from an academic professor and also from different perspective of different innovation from even from students. Right. So to give you a good inspiration, you know, what kind of things that academia can bring in. And also then you back to industry again and say, I want to do something again, right?

And then, and eventually after a few years, you said, no, I got too much idea. You know, you have too many different things going on. You want to do a good, deep research. Then you became a professor, right? So then you back to industry again to do different things. So like a heat treatment, I sometimes see myself like a heat treatment.

You make metal, you're quenching and create a harness. Then you do tempering. You make yourself have a good stiffness, right? So, so that's the way I see myself. And I do believe in manufacturing is a noble profession. So you cannot just stay in one place. And for one set of knowledge and for a lifetime, right?

You need to have a good heat treat process, give you enough different type of strength, stiffness, hardness, eventually you have a balance of capabilities. That's what I'm saying, yeah. 

Natan: Yeah, it's amazing how you say, you know, manufacturing is a noble profession. It's so, it's inspiring because, you know, I always think about winemaking.

You know, that fundamentally, humanity made wines of all sorts and beers of all sorts, like, forever, and it's manufacturing. So much of it changed nothing and changed at the same time. Sometimes I think about it through the, we are the Homo Faber, you know, we're the, the technology wielding mammal, you know.

That's correct. Manufacturing is such a, an important facility of that. So, We both spend time thinking about like where manufacturing and specifically humans in manufacturing are going towards, especially with the, you know, there's no shortage of technology in manufacturing operations. And maybe before we dive into that and like provocating a little bit, you know, and kind of thinking through your recent published book, you'll tell us all about it in a little bit.

But before we do that, let's do a quick exercise. Where did you start your career in production when you were like, first production line? Take us there and like, what is your job? What are you doing? What are you making? How does the production line look like? 

Jay: Well, I start out from machine tool industry, right?

In early years and the NC programmer. You can't believe in 1983, NC programmer. We use a APT language, right? Automatic Programming Tool. We call it APT. Then you punch a tape. Yeah. A tape, then you bring to the post processing machine. No G code? No G code? No. You load it to the machine based on different type of configuration.

You could like a Makino, Yibang Makino, and Warner Swissey, and FANUC. Different machines. So you have to reconfigure them, right? So basically you have to program, Put the blueprint on the table, you program like a cutter, tool one. 

Natan: And the blueprint is of the part. It's a part. The blueprint is actually somebody took an isometric drawing.

Yes. In hand. Yes. From a board on transparency. And that's what you're programming. No CAD. Then 

Jay: you have to make a circle. Says C one. Yeah, line L one. . Yeah. So you cutter go left. C one, then two L one, then turn right, turn left. So it's, this is called a PT tool, right? Amazing. You build this tool, then make a tape.

You go to the machine, load it into the post processor and the controller, then you control the cutting tool. But that is numerical control, which is the foundation for today's automation system, right? Yeah. So later on, we had the opportunity. Work with a, early years, a automotive industry, General Motors, when I worked for robotics vision system, RVSI in Long Island, New York, invested by General Motors to build a 3D vision guided robots to make a welding and the sealant, automotive assembly.

So the question will be, that was early years of automation. So people have a very vague idea how to do them right. We were challenged by Japanese auto industry, right? So, hey, auto industry Detroit was a big three. So, hey, we got to do better in quality, right? So we have Demings or whatever, Toyota production system, but that's not just about learning, it's about doing.

How do you make a better quality of tools, robots, right? And also machine. So, but how to do it? Not just machine vision, but also fixture, like car body, when you move, when they start settle. But they may shift, but you pre programmed robot, but if body shift, what do you do? Ah, so you have to adjust robot path automatically.

That was a high tech, 1984. Right? You can adjust the pre programmed path dynamically, right? Every car. So you can continue the welding without any interruption. So if 

Natan: you think about yourself back then, you know, there's no internet, there's no cloud, there's no all these kind of things, like, what are you thinking back then is the, whatever, if this would happen, this is amazing.

Jay: Well, see, every generation, we have a different type of a game changer, right? Yeah. 1960 was the birth of robotics, robots, Joe Engelberger, right? And uh, Unimation, 70 and the late 60s, 70s, Japanese took it as a major kind of opportunity to build robot, to use robot. Also 

Natan: 70s is like PLC is coming for real, like late 70s, yeah.

Jay: Yeah. Because like 1971, 72, the oil crisis, right? So Fujitsu had a major change. You say, Oh, I got to do some electrical stuff. And that's why FANUC came about. The FANUC name called Fujitsu Automatic Numerical Control. That's what the name came from. FANUC. Yeah, that's right. After oil crisis 1972. Right? Yeah.

I didn't know that. That's so interesting. Yeah, that's exactly The name came from after, no. 

Natan: The name I knew, but the correlation to the oil crisis and that's what pushed 'cause oil crisis. Yeah. 

Jay: Hydraulic system. You wanna change to electrical systems is what it's, right. So what, what I'm saying that the US we, we got, we were challenged by the automobile cells in the US so we say we have to make a better car.

Right? So at that time, the change a game changer is robots. Mm-Hmm. and quality. What's a problem? So we're competing with the Japanese vehicle. Now, how do we solve the problem? Using automation. What's automation? Robots, right? So robots are a machine, right? So it's a numerical control machine. But Jake, does that 

Natan: mean we kind of, also our view is like mostly a decade, like the horizon, you know, like how far we can see with human vision?

Jay: Well, in fact, when people thought about robot, right, and one of the big thinking, one, in fact, during the 1980s, Uh, we had a lot of people talking about, of course, people worry about job, they're pretty spied robots, right? They also, people say, oh, one day they'll become a lights out factory. Right. No people.

Yeah. But that's almost impossible. But people was worry about making a story, right? Right. 

Natan: Robots taking over the world and taking our jobs. 

Jay: Yeah. But when you physically do it, you found out it's not just about robots. So we had a called Manufacturing Automation Protocol called MAP, MAP. That's basically called today's Industrial Internet, right?

At that time we called Manufacturing Automation Protocol, MAP, right? So that's today's Industrial Internet. Exactly the 

Natan: prerequisite, right? Yeah, this is the amazing part. There's like, you know, human problem kind of mapped to time and space. It's like, it's the same problem. They don't really change. Like you're just trying to optimize them.

Yeah. Which brings me, not exactly to today, but In 2020, you published a book titled Industrial AI, Application with Sustainable Performance, and this is predating machine learning. And of course, in the research realm, AI is widely discussed, but like well before all the madness that we're seeing now with generative AI.

LLMs and the like and the maturity and what's the premise of the book and what's your perspective four years later so we can like get warmed up into the discussion of the hype cycle and what's the biggest opportunities of the stuff as it comes to industrial applications. 

Jay: Well, let's first of all talk about motivation, why that book came about.

Natan: Yes. And also about the timing, because I think it's very interesting, the timing. 

Jay: Yeah, yeah. So during my transition sabbatical with FastCom, FastCom is a global manufacturing company, right? So that book started actually 2018, by the way. 2018 was writing the book, and that's based on what we have learned for the last 18 years during the NSF Intelligent Maintenance Systems Center.

We've been building, using data from different type of machines, how to. predict the maintenance, how to predict the failure. So we thought it may be a good time to compile what we've done for the last 18 years before NSF funding is over for total 19 years. That was a planning to make it like a kind of a conclusion what we've done for 18 years research.

Of course, when the opportunity came, I was kind of convinced by Vascon, say, hey, why don't you help us to build the future global manufacturing program as a vice chairman and board member? So during that time, I said, well, The biggest challenge is not about technology, AI, it's about talents, okay? How you train the people at a large scale.

I'm not talking about training 200 people, I'm talking about 10, 000 

Natan: people. 100, 000 people. 100, 000 people. We need 2 3 million just in the U. S. alone. Yes, yes. And without semiconductor, now we need more because, you know, Exactly. ChipHack is bringing back factory. We need more. 

Jay: So at that time we had, I call the three S challenges.

First S is a scale. You want to do big scale training. Not one. It's a thousand, 10, 000. Second, speed. You can't just learn AIs like ad hoc. Oh, I learned some basic programming. I take course from the, you know, Coursera course and one day, two day, three days. Okay. You have to develop AI, test it, implement it in two days.

Right. We did some example in one month. Scale, speed. Yeah. And systematically. Systematically. Yes. Because traditional AI machine learning is that you and I together, let's say two of us, we have the same data sets, same problem to solve. We each of us using the same data set, same algorithms, but very often two different results, right?

You say, why? Well, just like a chef, two chefs, same kitchen, same 

Natan: fish, right? I know the mathematical, you know, computer science answer to that. It involves fuzzy logic, but you know, it's, uh 

Jay: Because the feature selection is different, okay? But I'm a good chef. When you cook, The way I put a salt, you put, you turn around, you cook thing very different.

Even I give you same fish to different shape, different taste, right? Different taste. So same thing for AI. So if you do that, there's a big problem because In engineering, we normally call system engineering. If there's a crack in your bearing, in your turbine blades, it's a crack. You cannot say one person say maybe a crack.

Another person say, well, likely the crack. So you got to have a consistent way to draw decision, right? So I said, well, how can we develop something so I can train engineers with 3S? Speed, scale, and systematically. That was the whole motivation about this book, actually. So give us 

Natan: an example of what is a great AI application that has sustainable performance, that can be mapped and scaled.

Jay: For example, Uh, we work with Toyota, right? Georgetown, Kentucky plant, where a Camry model was built, right? 2006, we had the opportunity to work with the president of Toyota, Georgetown, Kentucky, which is Mr. Niimi san. Mr. Niimi was the president. So he said, we have some random failure of production, including compressors.

Is there any way we can remove those failure out? Right? Because it's random. Exactly. I don't want to have one year happen in January, another year happen in February. They randomly happen at different times. You 

Natan: can't reproduce only in a car. Yeah, 

Jay: so can we do that? So we took that as a very great challenge and we implement a principal component analysis and a support vector machine.

And ever since 2006 until last night, no failure. So for almost 18 years, no failure. Well, there's no surprise to downtime, but there's a maintenance downtime, but I physically schedule you to down for maintenance. 

Natan: That's fine. But how did the support vector machine catch this in the beginning when it's just ramping up and collecting 

Jay: data?

Well, it's a classification tool, right? How do you classify the search? When machine is a search, compressor search happen, the machine. Just shut down in two seconds. But how do you prevent a surge? By using the flow sensor, by using the pressure, you can predict and classify potential risk of a surge. 

Natan: So you created a feedback loop, studied the pattern and the feature, classified, and then created action drivers on 

Jay: that.

Yeah. So what you do is if you detect the likelihood of surge, You can open the inlet guide valve earlier, relieve the pressure, so that the machine never have a sudden failure, 

Natan: right? That's a great story. Now, can you imagine implementing what you did in 06 with what you have today, Cloud, Industrial IoT, Gen AI?

Now go back to that point and like, how long did it take you? 

Jay: Well, that project took us six months, right? We had mistakes first, we located the wrong problems. But later on, we have a number of other projects, sometimes three months, sometimes one month, so eventually my question will be, if the problem need to be similar problems, I don't need six months, I can finish in three weeks, three days, how about three hours, right?

One day, right? Yeah, so that's my 

Natan: question. If the same problem from 06 to today, you know, so it's a proportion of almost like 20 years, right? How long would it take today? Days. We just got days. You're saying we're in days. So we moved 

Jay: from months. We'll be able to finish in two to three days, which will really demonstrate some of the problems we did this last year, did in two and a half days for the special project of a space shuttle from Japan, NASA.

They can detect the likelihood of failure. That was a data challenge project we did. So students spent two and a half days. So now 

Natan: is a good introspection to the premise of the book and the impact and the great example and we see the progression from technology side, but let's talk about the human aspect for a second, you know, this is sort of where I'm very passionate about and this idea that lean needs to be augmented, hence augmented lean and the approach that, you know, you have to, I'll tell you this like for my, call it like professional paranoia, okay?

I kind of feel that in the traditional sense, digital transformation as we knew it, you know, the 25 years or so of PC revolution, you know, large scale office suites, uh, cloud, you know, everything that drives our businesses, forget operation for a second, if that kind of skipped over the people in the factory line that were still filling up, uh, you know, Kanban cards and whiteboards, we used to call that industry 4.

0 for a decade. Now, I think less so, because it's kind of done in a way, we can circle back on that. But my point is, if the same operation people, you know, I'm speaking about the associates building stuff, the engineering kind of stuff. are not getting access to this AI stuff, Gen AI and, and deep learning and whatever, then we're missing something here.

How do we make sure that the people are, stay there and involved? Let's discuss this a little 

Jay: bit. Sure. Well, AI, right, to me, Augmented Intelligence is the basic stuff. Forget about Artificial Intelligence. First should be Augmented Intelligence, right? Yeah, 

Natan: that's what my PhD advisor, Patti Maas, to say, Think about it.

AI, think about IA, which is intelligence augmentation, you know. 

Jay: Yes. Yeah. So from augmented intelligence, which you mean, how do you help the people to do things they could not do, or they may not be able to see the problems. So I call it visible and invisible. Human being, we grew our knowledge based on the physical, evident experience, like we made mistakes.

Natan: What the senses can 

Jay: sense. You got burnt, you remember, right? But how about those things you have not encountered? Sometimes we see the problem, we didn't know, even know why. So I think that one of the things that, because AI is a data centric system, you have to have data to drive that. Of course, there's also physics informed AI.

You can integrate data in the model. But again, the purpose to help people to see things you don't see. For example, wind turbine, you have ice accumulate on the blades. So, oh, your power generation dropped. But why? Ah, because in the mountain, you got a low temperature in the night. You got high humidity. Ah, that's why you got ice accumulate on the blades.

That's why you are losing energy, power generation. Turbine designer, they didn't know that one, right? Because wind farm owner can be owned anywhere. Right? It can be a mountain, it can be offshore, it can be along the highway, right? So eventually what I'm saying is that AI is a good way to argument what human beings cannot do or don't do well.

Of course, one day AI can help autonomous intelligence. That somehow, you don't want everything you have to, like repeating work. A human being has to repeat the same kind of work, right? But if you can use that, AI can autonomously repeat that work. So remove your worry out, right? With a worry reduction.

That will be the best, but not to replace people alone. Yeah. Basically, you reduce redundant work, that's what it is, okay? 

Natan: But we kind of touched on this, and I think sometimes people are like, almost afraid of talking about this, you know, the geopolitics of the global supply chain, you know? It's like, we're seeing the tensions between, you know, China and the US.

We're seeing the global supply chain dependency, right? And people are speaking about reshoring. And still, we live in this global economy. You know, the US is missing 2 3 million, Europe is missing 2 3 million, I don't know what's your view on China. It depends on like how you think about their evolution, but obviously huge economy.

What's your take on all that as it comes to like the role of AI? You mentioned before the robots have not taken over, in fact, you know, we still need the humans and like, how are we going to deal with this problem globally that we don't have enough people in operation given AI? Well, 

Jay: if you look back, right, 20 years ago, we have offshore, right, offshore, uh, people start globalizing many things, right?

Right. That's why in 2000, China had a great opportunity to outsource into lots of contract manufacturing. That's how it started, right? Right. But over years, um, even like semiconductor fabless. Right. If Intel even went to Fabless and many companies to Fabless. So eventually this is how Korea and Taiwan start making chips.

But after 20 some years, they become a giant, right? Right. So not one day transformation, many, many years, right? Now we say 20 years later, we say, Oh, we want to do more. Well, there are some numbers. In the U. S. we have about 150 million workforce But only about 15 million workers in manufacturing, but we could have more, but depends on what type of work you want more.

If you want a lot of low skill workforce, it is what you want, or you want to increase the GDP per capita. Okay. We're now 80, we 

Natan: want option number two. Yes. 

Jay: How do we do that? I want to see our GDP per capita is 100, 000. Not 80, 000. Okay. Singapore, Switzerland, today's exceeded 100, 000. Right. So if that's the case, so, Oh, what can I do to improve my efficiency and productivity?

That's the purpose, right? Yes, absolutely. Focus on replace people. That's not the problem. The problem is productivity, competitive position. That's what we want. So if that's the case, every country is a problem. Every country, China had a great growth for the last 20 years. Then because they leapfrogged many, many stages, now certainly you have a geopolitical constraint.

Now you have to re examine what needs to be redone. So that's why if you look at the global Lighthouse network, there are so many companies trying to do Lighthouse. Why? There are two reasons. Number one, the baseline they came from was okay, but not good enough, right? So baseline was much lower so they can easily improve 2x, sometimes 3x efficiency.

Natan: Right. You know, we're both on the same sort of lighthouse expert group. I should have checked that, but I think we've crossed already 50 out of 153 now. Yeah, but more than 50 are in China. 

Jay: Oh, yeah. China has over 50. Yeah, almost 

Natan: 60 now. And this is sort of the point I'm driving is like, you You mentioned, you know, 20 years is a long time and started small and outsourcing and contract manufacturing, but then it becomes, you know, I say from contract manufacturing to core competency.

Yeah. And when the core competency shifts from one place to another geographically, then you have a deficit where it's moved 

Jay: from. Well, it's about ecosystem, right? Just like Detroit used to be, had the strong ecosystem back in 1970s, 80s, right? So almost every company in Detroit, they are really, truly dovetailed together.

They are so integrated, they can make almost anything, right? Of course, for the last 20 years, you see other parts of the world that are doing similar things. So that's why you can see EV production, a very good supply chain. But we have to ask ourselves, what can we do to select the right area to increase our productivity, right?

If you want to do everything, that's not necessary. Because you're probably going to run out of energy to do everything. So for example, electronic manufacturing, we want a high end AI chips. You want to do HGPUs, future servers, stuff like that. Yeah. So you want to make those high end, we call high value.

Sometimes may not be high value. Okay. What I call high value system, satellite system, right? But you want a high value system, the reason is we don't have a high volume workforce, but we want to create a high value workforce using machine learning as a good productivity tool. Okay, 

Natan: this is important. So this is the transition from high volume to high value.

Yes. But geopolitical tensions aside, I'm still an optimist and believe in the human race in general. Maybe I'm too naive. But you brought together with Foxconn, I'd say, operations to the U. S. and research to the U. S. So we're seeing such a major company like Foxconn coming here and setting up. Can you share a little bit about that and like what was the genesis of the AI division for Foxconn and what's the goal and how are you playing in this ecosystem?

Jay: Yeah, FosCon had early establishments in US back in the 1980s, right. So there are so far, it's about 16 locations in the United States. I would say probably even more now. Wisconsin was the most recent one because of the one I physically was involved, was responsible because of the early stage of the year to look for the new Potential opportunity to build manufacturing, right?

So later on, we changed to a new area to make high end, high performance computing devices, right? So because it's high end and it's a high value, it's using a lot of automation. Right? SMT, which is a fundamental platform for Foscon to make a high value production. And if you see what's happening right now with all the GPUs going on, you know, worldwide with the A100, H100, and Foscon also play a big role in here, making the GPUs, right?

So Wisconsin has a great foundation to make a strong and very high value electronic system there. So to me, that's a new addition. But sometime you have to redesign the production. Because it may not work for all areas. We have to, I'm sure you have to redesign the system. So you have more integrated automation system, not fragment automation system.

And second thing, we have to start looking at suppliers that US has a very strong presence in domestically. We can support that. Even we have a new design. And to educate the new workforce, right? For example, we do an assembly system. We adopt a lot of high speed AOI, machine vision. That's where we build industrial AI there.

Because we have lots of data. We create a scheme called stream of quality. Yeah. So traditionally inspections, each individual station only inspect one area. Now we connect the quality data stream like a river. So you connect the river, I mean, you have to connect data flow together. So the AI have to work as a stream, not with individual station.

So a new skill sets that most industry. I would say most companies didn't have it, even today I don't think they have it, but of course we were leading 

Natan: that. So this leads us to all this development and if we think about the future, so, you know, when I think about the future, I think that if we miss the education piece of this, and this is going full circle to, you know, you talked about the cycles you went through and back to academia to focus deeply.

Yes. If we think about the future. What's the role of education here? What do we need to do as it comes to making sure that the next generation is ready, given all the transformation that we've seen? 

Jay: Well, we have to think this way, right? So I often use a 3T to conclude. One's technology, one's the tools.

What is the talent right? Technology continue evolving, but we need to have a tool to have a systematic way to do things. What technology bring about. Then we need a talent can use a tool to do the every day. So the three TR connected. So one, a big thing I believe right now the industry. Most companies say, well, how can I use the data?

I have a lot of data. I don't know how to select the right data for machine learning. I want to learn AI, but how do I implement AI? So what do 

Natan: we do? Like, what university programs need to evolve into? Like, what's the curriculum need to evolve into? Well, that's exactly, 

Jay: we created called the industrial AI curriculum at the University of Maryland.

And so we are creating a master of engineering. In industrial AI program. Okay, so with a four P system, right? Principle based learning, you learn courses, basic stuff you can learn by yourself or we teach you fundamental one first. Second one is the most important called practice based approach. Mm-Hmm. . So we use a real data from industry.

We've been accumulating over the years. Uh, I see. We give data dataset to students. Hey, take the datasets. Use the tool you learn, solve the problem. Because we had previously solved the problem, so we have the baseline to compare the benchmark. The third, project based. Hey, you have to go to find your own datasets.

Natan: Right, in a real manufacturing environment. In a real manufacturing environment. Yeah, 

Jay: perfect. Or from the open source. Right. But that has to be usable. That's useful. The fourth one is the professional base. You have to be able to be like a masked black belt. You can teach other people to do one, two, three, right?

So called 4P. Principle based, practice based, project based, and professional based. So this 4P approach to give us a very good edge to develop a program to train 10, 000 engineers. This is our goal. Train 10, 000 engineers in 10 years. So we're going to 

Natan: include those links to the program. It's online, right?

We can, we can include. Partial 

Jay: information is online, but I think we're releasing soon, 

Natan: yes. Okay, whatever you can send us, we would love to share it. We're contributing to this ecosystem. And you know, this is what I think this conversation is really about. And we talked about the 3S, the 3T, and now the 4Ps.

And we'll send a quiz, like, you know, after this. So, if Professor Li gives you an A plus or not, we'll see. But certainly this has been fascinating and I've learned a lot. It's always great to get together and really enjoy our conversation. We will do this again. Anytime. Yeah. I really appreciated you coming on the show and, uh, thank you so much.

Well, 

Jay: thank you for the opportunity. 

Natan: Perfect. Thanks a lot. 

Voiceover: Thank you for listening to this episode of the Augmented Ops podcast from Tulip Interfaces. We hope you found this week's episode informative and inspiring. You can find the show on LinkedIn and YouTube. Or at tulip. co slash podcast. If you enjoyed this episode, please leave us a rating or review on iTunes or wherever you listen to your podcasts until next time.