You want to know more about what it is we are doing at Luminovo? You know nothing (or very little) about the electronics industry? Or why we think it is exciting? This post is for you.
Not too long ago we decided to put all our eggs in one basket and build the operating system for the electronics industry. To be frank, not all of us grew up thinking: “I want to build the operating system for the electronics industry!”. Instead, many of us grew up thinking “What a nifty car!”, “What a dandy phone!” or “Why is my train always late?”. Turns out there are few industries more omnipresent in our day-to-day lives than the electronics industry. In fact, lots of human progress made in recent history can be traced back to technological progress involving electronics. At Luminovo, we want to contribute to accelerating that technological progress. That’s why, by now, it rolls off the tongue easily for us to say that we want to build the electronics operating system.
Before we explain what that last sentence actually means, let’s take a quick look behind the scenes of the electronics industry and the stakeholders and processes needed to bring an electronic product to life.
OEM stands for Original Equipment Manufacturer. They are the ones designing and marketing new electronic products. Some OEMs (like Apple, VW or Siemens) sell directly to the end-user and some (like MTU, ZF and also Siemens) offer components that are re-sold as part of a larger product by another company.
Many OEMs don’t do any manufacturing themselves. So, the name Manufacturer can be a little misleading.¹ This is where EMS come in.
Electronic Manufacturing Services. They are the unsung heroes of the electronics industry and the ones that actually manufacture the laptop, phone or tablet you are probably using to read this. They also build the satellites, cars, medical devices and household appliances (the list could go on) that keep our civilization afloat. 🚀
Some EMS not only manufacture but also design new products (but none of them market new products–that is OEM territory). Generally, many EMS are trying to differentiate through value-added services, such as testing, logistics and (as mentioned) helping with designing new products from the start.
💡 Bringing together design and manufacturing is actually key to shortening innovation cycles in the electronics industry. Read more on our take on this below.
Yup, no acronym lurking in the shadows here. The last stakeholders in the electronics industry you should be aware of are the distributors and manufacturers of electronic components like resistors, capacitors or integrated circuits. The OEMs sometimes directly collaborate with manufacturers (like Würth or Texas Instruments) during development and, if they have enough leverage, negotiate special prices for their project. However, there are (literally!) millions of different components that you could put on your PCB (Printed Circuit Board, you’re welcome). Distributors (like Farnell or Digi-Key) help EMS and OEMs get a grip on this cornucopia of electronic parts.
A Request for Quotation. When the OEM finished designing a prototype of their product, they contact one or multiple EMS to get a quotation for building the physical product. They usually want to know:
“How much is it going to cost me to build this prototype? How long is it going to take? And what can I expect if I enter series production later on?”
To answer these questions, at the EMS, they wade through the entire list of components needed to manufacture the product (called the Bill Of Materials, or — you guessed it — the BOM 💣) trying to find the best prices and lead times for the OEM. They also look at the design to understand whether there will be any manufacturability issues and estimate the labor (from both machines and humans) that goes into assembling the final product.
In the grand scheme of things, the RfQ process is just a small part of the entire value chain. But it is the one we decided to use as our entry point into the wonderland of electronics. 🌈
Before we tell you why we are starting with RfQs, let’s take a step back to appreciate how involved the process of developing electronic products really is. People say that good fences make good neighbors, but let’s take a peek at how developing a software product compares to developing an electronic one.
⚠️ Disclaimer: We will not be talking about what it takes to build something people actually want. A startup’s dreaded (or cherished) product-market fit. This, I can only imagine, is just as hard for software as it is for hardware.
To develop a software product you first and foremost need a software engineer. You will usually need some frontend and backend expertise. A little DevOps and if you are lucky someone with a knack for UI/UX design. On big teams these are different people, but you could get a product off the ground with just one person. To deploy your product, you spin up a server in the cloud. To test it you write some more code or click around to get instant feedback the minute you’ve finished writing your code.
To design an electronic product, you will need an electrical engineer, a PCB designer, a product designer and a mechanical engineer. To manufacture it, an industrial engineer and someone to negotiate and procure all the different parts on your BOM. And, of course, you will need people (and machines) to actually carry out the assembly of your product.
At the risk of waking Captain Obvious: the more people that need to coordinate the more difficult it becomes to collaborate effectively. To make things worse a lot of this collaboration needs to happen across company boundaries.³ OEMs. EMS. Distributors. Manufacturers. Remember?
This situation is not about to magically disappear. A full separation of concerns is not realistic for electronic products. The casing influences how much room you have for your PCB. The design of your PCB influences the effort needed during manufacturing. The choice of components influences cost and availability.
Now you might object that, in software, the choice of algorithms also influences your server costs. Collaboration between your product manager, UX designer and engineers is also important. But the stakes for hardware development are higher. Few software startups fail because choosing the wrong algorithm made their unit economics unsustainable. Original software manufacturers do not need to worry about their supply chain. They don’t get unlucky and have to choose between redesigning their software or having to wait for 6 months until the component they needed for their prototype is available again. And fixing failing tests by iterating on your design is not a matter of changing a few lines of code. It means re-involving all of the stakeholders mentioned above.
We are not saying that software is easy. But hardware is harder.
That’s where we come in.
So what to do about this? For the lack of a better alternative, here’s a bad comparison. Luminovo’s Electronics Operating System is for Electronics what GitLab is for Software. GitLab lets you track and version your work and collaborate on resolving issues. It gives you continuous feedback on your “designs” and lets you deploy them into production with a mouse click. The three tenets of our operating system are collaborative, data-driven and automated. The analogy does not work as well for the second one, but you get the point. Wouldn’t it be cool, if developing electronics was more like that? Wouldn’t it be cool if we helped speed up the development of autonomous cars, spaceships to Mars and life-saving medical devices all at the same time?
It will be a long and winding road to make that vision a reality, but it also took seven books to defeat Voldemort. And we are here to stay.
Our first stop on that road is building a cloud-native application to connect and support OEMs and EMS during an RfQ. By now you can probably guess why the RfQ–a central touch point between the two major stakeholders in the electronics industry–is the natural starting point for our electronics operating system.
Today this process is dominated by the use and abuse of Excel, data siloed in on-premise solutions and the ensuing friction of trying to communicate between these silos.
“Any organization that designs a system (defined broadly) will produce a design whose structure is a copy of the organization’s communication structure.” — Melvin E. Conway
In an industry where collaboration is paramount and design and manufacturing are inexorably intertwined, Conway’s law makes for a gloomy forecast given the current state of affairs.
Luminovo to the rescue. 🚀
• • •
If you want to learn more, check out our website. We are always looking for new Luminerds to join the family. 🤗
Thanks to Erin Bacsy, Benedikt Ummen, Magdalena Witty, Sebastian Schaal, Sebastian Schuon, Bernice Ruban, Benoit Miquel, Benedikt Bünz and Lars Thorben Neustock for your feedback. ❤️
• • •
¹ Most OEMs don’t do manufacturing, but some do. To add to the confusion, through what must have been a series of unfortunate events, the first example on the Wikipedia page for OEMs is Foxconn, which, as it turns out, is the largest EMS worldwide.
Seeing that we are already in the footnotes and have time to speak about more tangential things, a quick word about geography: Contrary to popular belief not all electronics are Made in China! It is true for most consumer electronics, but many products of other industries (such as automotive, medical, industrial or railway) are still manufactured in Europe (or the US). The more you know!
² Image credit goes to Ghania Riaz. Thanks! The icons are all from Flaticon.com, courtesy of Freepik, Vectors Market, Icongeek26, surang, Dighital and Becris.
³ There is a long history of academic research, pioneered by Ronald Coase and his essay The Nature of the Firm, addressing the question of when and why company boundaries form.
“A fundamental tenet of capitalism is that the invisible hand of market competition is superior to that of managerial oversight as a coordinating mechanism between actors in a market.” – Clayton Christensen, The Innovator’s Solution
When you are able to specify, measure and predict the important interdependencies across an interface between two companies, collaboration becomes tractable. It is only when there is insufficient information across company boundaries that market failures occur. It gets “too expensive to negotiate […] between otherwise independent parties” (ibid.), and managerial oversight becomes the superior coordinating mechanism.
In the electronics industry specifying the interdependencies (like lead time, cost and manufacturability) is easy. It is their measurability and predictability that we want to improve.