del20nd wrote: ↑13 Apr 2023, 04:22
Ellipse wrote: ↑13 Apr 2023, 00:35
This is an interesting discussion. del20nd you mentioned your experience in hardware manufacturing? Anything (details, best practices, advice, etc.) you are able to share, if it is appropriate?
Sure! I can talk about the subject generally, without going into the nitty-gritty specifics of how the company was run. (That information probably wouldn't help too much anyway, as they worked on a very different scale than this project. In fact, to that end, watching your google talk about the small scale you're working on was fascinating because of how it compared to larger scale manufacturing.)
Quick background for perspective on where I'm coming from:
My experience in manufacturing comes from four years of recently developing for, supporting and maintaining the ERP system of a medium scale US based appliance manufacturer in the role of a software analyst. The platform their ERP ran on was an IBM AS/400 / iSeries / Power System. In other words, the platform that 122 key Model F/M keyboards are actually intended to be used on
. I've moved on from this particular company, but am still generally working on the IBM platform.
For those that don't know, an ERP (Enterprise Resource Planning) system is more-or-less the backbone software package that any large corporation uses to track their business. Distribution centers, manufacturers, insurance companies, grocery store chains, etc all use ERPs that are tailored to their specific business needs. There's leeway in how these get used from company to company (some are allot more paper oriented than others), and they range from being out-of-the-box installs to extremely customized. In extremely generalized terms, a manufacturer might use an ERP to track Bills of Material, Material Orders, Finished goods, payroll, resource planning, shipments, invoices, audit history, etc, etc, etc. Which of those things, and in what capacity they're used, will vary wildly from one business to another.
Given that the ERP touches basically everything in the company, that I went out of my way to learn how the business was run (because I'm a bit of a knowledge sponge), and I happened to be employed during one of the most frantic, disruptive periods in modern manufacturing history, which required allot of out-of-the-box thinking to keep operations moving along, I have a decent thousand yard view of how it all comes together.
So anyway...
Tips and stuff:
I guess in general, given that it's such a completely different ecosystem, be wary of tips from anybody working in medium to large scale manufacturing
. You already know this, but the general knowledge around making 1,000 of something is wildly, wildly different than the general knowledge around making a million of something. In seeing how you overcame obstacles, I was starting to wonder, for example, if scale played into why you went for aluminum cast casings instead of plastic for the new keyboards, even though nobody would've batted an eye at plastic since it's "period correct" for the newest project: I started looking into it and it seemed like maybe the die casts for aluminum at a small scale, combined with the higher labor cost made more economic sense for you than investing in the very expensive injection molds that would be required to make plastic parts, even though there'd be a labor savings. Or maybe, if it didn't save money directly, if the cost difference was small enough to justify the higher quality case.
I don't need to tell you that your keyboard shouldn't be managed the same way as a production line churning out 6 million CD players of course, but I do see allot of questions like "why is this keyboard $420 when it doesn't even come with X Y Z premium super-premium 25 year warranties that [really big manufacturer] offers," and I think this is just people not realizing how much scale plays into the pricing we see on electronics today, and how spoiled we all are by that. These keyboards cannot ever have the same economy of scale as Logitech, that the margins on these boards may actually be lower than what a big manufacturer pulls, and you have to be realistic about what things cost when there's one person working on making only a few thousand of them!
If I try to think of something that could maybe broadly apply, while it wouldn't make a lick of sense for you to buy into a major ERP platform given your scale, especially because you're the only one touching the data, it might make sense for you to come up with data management conventions that mimicked the structure of an ERP if you haven't already.
I'm not sure how you're doing this now, but at the absolute minimum I'd have a religiously kept system to track incoming parts, bills of material to assemble orders, material orders for what's required for each variant of keyboard assembly, finished goods, invoices, etc. This organization is key to how larger manufacturers are able to turn around product. Keeping that info well tracked would definitely be helpful if you ever scaled up how much you're making. Given that I didn't work at such a small scale, I don't know what's on the market in this regard, or if maybe you're just really, really good at keeping Excel spreadsheets in order, but the big guys who are growing are very careful/organized with their data.
I'm probably not too useful on how to manage your supply chain; I did work with those guys but didn't do their job. I can say that it's a very quick moving, adaptive field, especially in the last few years. I guess I could apply that to your situation by saying to be creative in how you overcome bottlenecks? For example, if your supplier suddenly decided that they're not going to powder coat the cast aluminum keyboard casings, don't rule out doing something like importing them unpainted and contracting somebody local, even an auto body shop. As long as you keep those processes well documented so you know what works and what doesn't. Supply setbacks are common, especially these days, but there are sometimes ways around them if you're creative, especially if you're the first one to think of the solution and can negotiate a price under the radar.
Another thing I can't really talk about is how to deal with overseas manufacturers, since our factories were domestic, and made particular effort to keep as much of the supply chain domestic as possible. We did have to contract out some parts where no domestic options existed (broad example, there's not one manufacturer in the US who makes electrolytic capacitors, so if you're a company that needs some of those you'll be working with an international supplier) but unfortunately all of my advice here is broad platitudes which are well known: be careful of your suppliers, as Chinese factories are notorious for ripping off patented designs. Given your scale, and the fact that you're not patenting anything, you might not have to worry about things like this.
***
That's all I can think off the top of my head.
Given what you've done here, you know way more than I do, but I'm smart enough to at least understand that I. I can make educated guesses about why you may or may not have made this-or-that choice, and can tediously explain to people in the Ars Technica comments section that no, you're not Samsung and no, you're not "ripping people off", that this is just what it costs to bring a few thousand Models F into the world, that it's actually a miracle that these things can be brought to maket so affordably at all and he's keeping prices this
low by not having the overhead of multi-year warranties and 24/7 on-call support hotlines, that if you can't accept that small projects will inevitably cost more money then you're not in the market for one of these keyboards, please just buy a mass market mech board if you want to spend $100 or less, etc, etc....
If I do come up with anything else I'll try to remember to post it, and if you have any specific questions that maybe you want to try to jog my memory on, please feel free to PM or just reply here.