Interview with Everett Frank of DigiSource Inc. on Where the Supply Chains should Improve

Hello everyone, my name is Ian Peterman, host of the Conscious Design Podcast, and with me today we have Everett Frank, CEO of DigiSource and a supply chain expert in electronics. Welcome to the show.

Thank you very much. Thanks for having me on, Ian.

Yeah, I'm really excited to have you on. We've Talked a lot about supply chain, especially and what's going on right now. It's a bit of a topic for everyone. We've struggled through and are still seeing impact of what supply chain means. And you've worked in supply chain for a long time. You've gotten back into it recently and doing some last chance hunting for some products.

Yeah.

So can you Share a little bit about, you know, your experience, supply chain, and what it's meant, you know, for your own process and what you do for people.

Yeah, yeah. Well, I mean, I guess one of the first things I would say is I've never seen anything like this, you know, and I started, I'm not gonna count the years, but I started in the mid-'80s, early '80s. And I can tell you that everybody that I talk to that, You know, with that kind of seniority, says the same thing that we've just never seen. You know, back then we had really vicious cycles. You know, we had about 18-month cycles where we'd go through these, you know, just slammed and then down and up, and we everything would be on allocation and da da da da. And you know, it was really difficult. And and this is worse. You know, it's it's just. I haven't just ever seen anything quite like this, quite so broad, and the amount of time that's going on. And so yeah, for anyone who is struggling with it, it's an extraordinary time.

Yeah, and one of the things we— I think we talked about a little bit, but is really important is, you know, the relationships we have with suppliers. And, you know, that's historically speaking, as long as you had a really good relationship with your supplier, you could weather something like this. But even now, that's, while it's super important, there's shortages and things like that. And what is your, you know, you have companies that come to you and say, we can't find anything. What is your kind of process? Is it— Yeah. Is it finding new relationships? Is it old relationships? Like, what is even a possible option when you do hit that final mark of, well, nobody is publicly saying that we have this part out there? You know, are you seeing more redesign work into totally shifting a product over, or what's kind of— what are the the directions you're seeing people go in order to solve, you know, this huge supply chain issue we're all having?

Yeah, yeah. Well, I mean, in a very broad way, the small to mid-sized companies, it's easier for them to execute a design change. You know, they don't tend to have the process overhead and the internal barriers to changing the product. So they can tend to be, you know, more nimble in that way. So I think that the idea that we're gonna design around that problem is more prevalent in small to mid-sized companies. In bigger companies, it's much, much more difficult to just sort of execute, you know, certainly a circuit change. I mean, in many cases, even qualifying a new component is a months-long process at large companies.

Yeah.

And then if you're starting to talk about going and changing the circuit itself, it's almost unthinkably complex for big companies. And so they have to go find the parts. And then with respect to how to do it, I guess you're right. It is sort of the— it's a time of relationships for sure. And knowing who to know may be the way to put it. You know, so it's a combination of people that you already know and people that they introduce you to. It's kind of all these things in terms of how do you go and find the parts. And, you know, I mean, I guess my advice to supply chain people would be cast a wide net, you know, or have somebody who knows how to do that and is doing that for you either way. But it's— I'm trying to think of the right word. It's nichey to find the parts, you know what I mean? It's very lucky almost. I'm not finding the right word, but somebody will come across the part and you got to find that somebody who knows where the parts are. And it's, yeah, it's just a big challenge. I think another implication is, I don't know if it's the death of just-in-time delivery, but it's certainly mortally wounding it. So that's another issue here is that JIT relies on short supply chains. And, or excellent forecasting. And now, how do you reasonably run an operation in a JIT configuration when nobody knows if you get the parts? And so that's kind of changing perspective. And I think certainly another impact is the nearshoring, you might call it, right? The availability of components nearer to your geography, certainly much greater awareness of that, and all those international risks, or there's a much greater awareness of that.

Right. Well, yeah, we're seeing a lot of impacts that we may have been able to ignore before, but now we can't. Stepping back a little bit, To, you know, assuming that we can have a supply chain that functions. Building out a supply chain is something a lot of companies, even designers, don't even know how to do. It's a chore and a job in and of itself. And I wanted to talk briefly about some of the compliance part of this because that's something, you know, we all like to design something and figure it out, but There's many layers to sourcing products and components in a smart way. And I know you've worked in that space, so, and you have software that you're working with to create a solution to be able to actually build a smart supply chain. Can you share a little bit about your experience and things that you look for? in building out a supply chain? What are the pitfalls that you can quite easily fall into when sourcing that we don't always think about the first time, but then somebody from compliance says, nope, you have to jump over here and that hoop is not open for you today?

Yeah, that's right. Yeah. Yeah. Well, that's a wide-ranging question. I think starting from the beginning, the compliance is largely an issue of what gets specced into the product, and it's a little bit less a sourcing problem. It'll lead to a sourcing problem, but getting the compliance correct up front and becoming aware of your compliance obligations is one of the most difficult things, and compliance is only getting more and more complex. Probably the best example is the REACH requirements. Everybody's pretty familiar now. We're all comfortable with RoHS and the unleaded, RoHS with the unleaded requirement. But now in Europe, there's a REACH requirement which has to do with certain substances of high concern, and it ultimately relates to how that product is disposed of in the future. So they're trying to track all this, and if your product by overall weight contains a certain threshold percentage of these unwanted chemicals, you are required to register that product in what's called the SCIP database in order to market that product in Europe and be in compliance with the regulatory requirements. And so this presents a challenge. Then you need to be able to figure out exactly how much weight of these obscure chemicals as a percentage is in my product, right, in order to trigger the, the skip requirement, right? And so, you know, these are like crazy. I mean, you know, 10 years ago, it's unimaginable we would be trying to track and trace these things to an individual product level, but that's where we're at. And, you know, similarly, if your product— if you intend to sell your product in China, the RoHS requirements are quite different in China than they are in Europe. It's much, much more limited actually in China because China doesn't allow exceptions to the requirement in the way that the European RoHS does. And so it's quite different. And so there's a lot of regulatory issues now, and it's only getting worse. So what we're doing then is trying to provide what we call a product risk management tool that assesses all these things. So we, you know, starting with, are the part numbers correct? Are they purchasable part numbers? What is the availability lead time and reporting all these sort of very basic metrics as well as all the compliance data and then giving a convenient way to cross-reference to alternate parts and some tools to help make those selections. You know what, like on a capacitor, you might have 15 different part numbers, right? There's their direct crosses, which is the best one?

Right.

And, you know, so helping to kind of solve, you know, give guidance to those as well as providing consolidated feedback, I guess I would say, on the design side. So we can assess the printed circuit board design, the fabrication design itself, and we can assess the assembly as well. So we use Valor, which is very popular software for that, but we consolidate the output of Valor into a readable easy-to-use format. And we provide all this for a few hundred bucks. So it's really, it's an unprecedented value to me anyway. It's an unprecedented value in assessing your overall risk profile of a product.

Yeah, well, it's something that your risk profile is a great, great word that a lot of, you know, I work with a lot of startups and they don't always look at that, right? You're in a fast growth. We got to get this out. It needs to, you know, start making money, otherwise people aren't going to look at it. But there's, there's this component of risk that, you know, shouldn't— in my, in my opinion, shouldn't be ignored. It should be— it's worth the time and effort to use a system like that. It goes, okay, well, here's, here's the components you should be looking at. Here's why. Here's why this component doesn't Doesn't meet these requirements. And, and actually have a picture because it's, like you said, it's getting harder and harder with regulatory requirements. And, you know, and some of them are, are really great. Like, we probably should be watching out, uh, certain chemicals, how, how much we put into a product. Uh, but it is that extra layer of, well, how do you track that? And I was very surprised that it's a percentage of weight, because some of that stuff, I'm not sure how you can even measure that accurately. So also, you know, the whole, well, how do you, how do you make that actually work in a smart, repeatable way that you could validate? So yeah, I think I'm gonna have to go look and see how they actually measure that.

Yeah, well, we can help you with that. So, yeah, so our tool reports it so we can go back and tell you what your, you know, what your weight is. Astonishingly enough, I mean, to me, it's amazing that we can even access that data and report it.

Right.

But yeah, it's insane. But we can, and we can report it. And, you know, I think too that, you know, electrical engineers are trying to do something amazing already, you know, with the products that they're designing and trying to create all these, you know, amazing things. And that's really where their competency should be. It shouldn't be in trying to figure out what's the weight of some obscure chemicals that are in the products, right? It's like, that's not where that expertise should lie. You know, the core competence of the engineers needs to be designing amazing products, and we need to support them with these tools. that makes them aware of these little landmines, but their focus needs to remain on being great at design.

Yeah, it's even more than ever before, it's a team effort. The idea of, you know, somebody going and building a commercial-ready product in a 1, 2, even 3-person team is kind of In my opinion, it's kind of dead. Like, you can—

It's quaint.

Yes, quaint. That was a quaint idea. It's, uh, much— there's so much more to, to think about and to put together. And it's, it's great. It's, as you said, like, you know, electrical engineers, they can actually focus, should be focused on the design and develop that idea. And then, then you have compliance, and then you have, you You know, import, export, and then you have all of these other areas that are supply chain. And I think we've kind of all hidden a little bit, at least in the design world. We've been like, oh well, we'll figure that out later.

Yeah, right.

Right? Like, look at my beautiful product. It's amazing. And then we kind of just walk away and start the next one. I think that kind of also needs to become a quaint idea. Not that we should be an expert, but we should be aware that there is this whole other side to supply chain and to compliance and bring that in earlier because if you design a beautiful thing, but nobody can build it because the way you designed it requires a bunch of components that— You know, Europe won't let you import. You're kind of stuck.

Are you really happy with that?

Yeah.

Yeah. And I think it's interesting. Everybody struggles with this sort of dynamic, right? Because the engineers want to, you know, design beautiful things and they want to sort of throw over the wall supply chain issues. Right. And like, is that, I mean, really even just starting with lead time and availability.

Oh yes. We're on it now.

Like, I don't know anything about that. That's somebody else's problem, you know, is a metaphorical attitude of engineering. And, you know, that's no good either, because you can't, as you say, if you can't get the parts and not compliant, and you can't build it, that doesn't do you any good. On the other hand, I don't think that the bigger stick should pass to supply chain or operations, at least in most businesses. In most businesses, the core competency and the, you know, the reason that they're, that they're, that they're going to be great or are great is because of their engineering. And the engineering needs to be supported and assisted with these things, not beaten into submission.

You know what I mean? I agree.

Yeah. So it really needs to be, I think, more about getting the tools to engineers and they need, they need to be willing to sort of use that information to make intelligent choices. But for God's sake, they can't spend their day doing that, you know?

No, they wouldn't design anything.

They wouldn't design anything, right. Where would we all be, right?

We'd have a very compliant nothing city.

Right, exactly. Right. But we're perfectly compliant, no risk.

Yeah. The perfect product, zero risk, zero time.

Right, zero time, yeah, yeah.

And so what else, you know, there's— so we talked about, you know, compliance and things like that. What else are you looking at? I mean, you mentioned, you know, design, the actual, you know, not even just component compliance, but the design itself, you know, making sure that it actually meets standards and requirements. What else in your experience do you, you know, engineering designs this great thing, it works well, now it comes time to manufacture it and build a supply chain around it. What other areas do you feel like should be paid attention to a little bit more? And, you know, not that obviously it's not the engineer's job to know everything, but What other supportive things should be present and we should be aware of, Wimber, when we are designing those products in order to make sure that the supply chain isn't going to go down a rabbit hole no one can return from?

Yeah, yeah, yeah. I think from the design side, I think that starting with the design components itself, the printed circuit board design, You should run Valor on it and use an intelligent database, something that can output ODB++, which more and more of the tools can do. Almost all of them. I think KiCad still doesn't do it. But, you know, there's a popularity, particularly amongst cash-strapped startups, to use tools like that that don't output ODB++ files or similar intelligent file systems that can then feed into Valor for simulation. Less so in the PCB, it's less critical on the printed circuit board Valor. And all PCB fabrication facilities use Valor before they run, almost all anyway. So you get feedback from the printed circuit board manufacturer, typically, you know, TQs, this kind of thing. So you do get feedback that way on that. On the assembly side, it's unusual really for customers to run a Valor simulation prior to assembly, which is a separate software tool. And we would definitely recommend running that. It's, you know, a few hundred dollars, you know, $350, $500, something depending upon the complexity. to run a Valor simulation and it tells you all kinds of things, hundreds of things get confirmed. And it's really, you know, I just can't say enough, the proper way to bring a product to market is use a tool that does ODB++ and run the Valor on both the assembly and the PCB. And in terms of the supply chain side, I would kind of say probably the biggest things I would highlight are number one, verify the part numbers. Use a tool that verifies the part numbers. A BOM scrub is typically called. You know, it's— there's lots of tools out there that will give you feedback on that. Ours certainly does it. But again, you find out if you've got a, you know, a slash in the wrong part in the wrong place, And then you, if you put, you know, some stupid number in the wrong spot, that's absolutely meaningless from an engineer's standpoint. But when it gets to a buyer and they can't, they don't know what the part is and they can't, you know, they can't match it at a distributor. And now you got, you know, so verify your part numbers, you know, somewhere in that early process and engage with your critical component suppliers, especially if you're a startup. So really at any product that's gonna go to any volume, but it's especially true if you're a startup because you need to build a relationship with the right people at your key semiconductor suppliers so that you get the right support. And that starts certainly with technical support, but it'll quickly become pricing support and then it will become delivery support. And the more that you, you know, startups are going to have very volatile schedules, you know, they just all do, you know.

Yes.

And so if you don't have a good relationship in place with your key component suppliers, you know, and I would, so I absolutely encourage, you know, my clients to be doing that, you know, I'll introduce them to the semiconductor companies. You know, it's like you need to have this direct relationship. And so a good relationship with Avnet and Arrow does not substitute for a good relationship directly with your strategic suppliers. So those would be my comments about sort of key best practices early on.

Well, that's good, and I think that's like Great point of having a direct relationship because it's one of the things I always try to tell people is you want those relationships as soon as possible so that you can start going through the process because you never know what kind of feedback they might be able to give you on on what you're working on or be able to warn you of supply issues things like that and yeah volatile. Volatile timelines for startups means that information can make or break you.

So it's— Absolutely. It absolutely makes or breaks, especially startups. And especially, you know, nowadays, the IoT market in particular goes from, hey, we just designed this, to we need 10,000 of them next month, you know, and we need 100,000 of them next quarter. I mean, that's so— that's like common in the—

Right. That's not even— that's not even a hockey stick. It's just normal growth.

That's just normal growth in these IoT products. And, you know, where are those parts gonna come from, right? If somebody doesn't do it. And, you know, a lot of what you're seeing right now is that the typical start to finish from sand to product of a semiconductor is about 6 months. You know, it doesn't matter what you see as a published lead time. You know, what you're seeing as a published lead time is just reflecting what they expect to allocate in their pipeline to that product. That's what you're seeing when you see a lead time. The actual manufacturing lead time of most parts is around 6 months. You know, so if you're coming in out of the blue and you want 100,000 of anything that was unplanned in that 6-month window, You know, well, where's that going to come from? You know, you just can't—

Right.

You know, it's not possible to turn on a dime like that in semiconductor manufacturing. You need to be thinking that if you've got something that's going to, you know, jump into— and I guess the caveat would be you would want to think of that as a percentage of that product's volume. Right. So if you know that somebody, uh, that STMicro is making 50 million of this, of this controller a year, and you might need 100,000 of them, well, that might not be that big a deal, you know. But if you know that that product is about— is a million or 2 million a year, and suddenly you're a 5 or 10% jump in their production, now you got a problem.

Right. It's, it's how much of their production are you gonna And yeah, 5% is a lot for something that's 6 months out. That's—

Yeah, yeah, because you think about it, if you're in their shoes, you've got to plan 6 months out, right? Well, you're going to plan for what you reasonably think you're going to sell in that period, plus or minus something, right?

Right. Yeah, you'll have a buffer, but not, not enough to buffer 5% each month.

Usually not. Yeah, yeah. So yeah, if you're a calculable volume of, you know, of what you think that part is moving and, you know, you may have a spike demand for it, you should really have a relationship with the supplier so that they've got you on their vision for production. And another, which sort of brings to mind another advantage, is that if you have that relationship, And you get it at the right level, they'll start talking to you about their product roadmap and about what their next-gen products are going to be. And they'll, if they really think you're cool, they'll ask you for input on their next-gen products. So that's a kind of a sign that they really like you.

And you're hot stuff if you get that conversation.

Right. So yeah, if you're in that conversation, now you know a couple things. You know that you're pretty advanced in whatever you're thinking about, because they talk to everybody, you know, so they have a wide scheme.

Right.

And it also, you know, from the perspective of a semiconductor company, the OEM is like a pony that they wanna ride. They wanna ride it in the pony races, right? And they wanna get on winning ponies. They wanna identify winning ponies, 'cause they don't sell their products direct to consumers or direct to end users, right? They sell their products to successful companies who can design it in and sell it and buy a million of them, right? And they're trying to pick the best pony. That really is what they're doing when you're engaging at a high level with them. That's what they're doing. They're trying to decide if you're going to be a winning pony. And so it's another sign if they engage with you that they think that you're your technology or your approach or your access to the market, these different things are going to be winners. So it's validating if you see them doing that.

Yeah, that's a good point to think about too, is that, you know, your supply chain, the farther, farther back you get, the farther away they are from a customer and the less they care about the customer in any way. They care about whether you can actually sell your product because that's how they sell theirs to you. So it's, yeah, it becomes a little bit of a different negotiation and a conversation and different trust level and they look at different things than, you know, someone else might in the supply chain, what they count as valuable.

Yeah, yeah, yeah. And their interests are just different and if you can align to those interests like, you know, a microcontroller, Right? It could go in an IoT product, it could go in a car. Well, which market, as a semiconductor marketing executive, which market are you going to devote resources to? You have to make a decision, right?

Right. Which one's growing? Which one's going to help them stay alive?

Yep.

So, yeah, that's really great to think about. Yeah, I think This has been, this has been really great. I think there's a lot of information. If you're a startup, definitely need to be thinking about all of these things. Don't skip compliance. Don't, don't, don't skip it. Just don't, please. Uh, we'll save you, save you a lot of time and heartache later. Uh, and for people that are wanting to get ahold of you or, or take a look at your software, You know, where can, where can people find you and get a hold of you and, and hopefully, hopefully just get the software and be and make sure their stuff is, is good to go?

Yeah, and keep their risk profile down. Um, so we're at thedigisource.com and you can see our, our products and our things listed there.

Perfect. Alrighty. Well, thank you so much. I really appreciate you coming on to chat and to share. And for everyone out there, look up DigiSource. If you're in doing any electronics production, get it checked, make sure you're saving yourself some pain later. You just avoid that and do it now. Well, thank you.

Okay, thanks, Ian. I really appreciate it. I always enjoy talking to you. I always learn something new every time we talk, and I look forward to the next time.

Sounds good.