So I got my first Original Equipment Manufacturer (OEM) prototype made. If you’ve been following my blog, you know I’ve been working on an Internet-of-Things devices for the past few years as a side project. It’s been a ton of learning, and a very worthwhile undertaking overall. I’ve learned so much: woodworking, CNC, metalworking, welding, circuit design, circuit building, PCB prototyping, Python (for controller programs), interfacing with Bluetooth, mobile apps, product design, and finally how to find a manufacturer to turn your idea into a tangible product. This post covers some lessons learned on this last step.
So the first thing is to put together a product specification. This has to be as detailed as possible. Any ambiguities are going to cost money later to flush out. It’s probably not possible to avoid some iteration in design, but you definitely need to minimize changes, especially if you’re bootstrapping. Be sure you have every component and interface spec’d out, even the off-the-shelf stuff. Find suppliers and include their cutsheets in the design package to the manufacturer. Don’t assume you’ll be able to find a particular component later. Make sure it’s available now and the minimum order quantity (MOQ) is something you can live with.
The other part of the product specification is the regulatory requirements. There are a lot of regulatory requirements and as the business owner it is your responsibility to comply with all of them. The product compliance tool offered by Compliance Gate seems like a good place to start: https://www.compliancegate.com/offer/. Honestly I have not gotten my head fully around this part of the process yet.
The next this I want to mention is where to find manufacturers. I happen to be in Vietnam at the moment, so found a manufacturer here. I was planning to visit the Canton Fair in China but that was before covid hit. Canton Fair, and other fairs like it, along with Alibaba for the online option, are typically the recommended resources to find manufacturers, but I have found these more suitable for Original Design Manufacturers (ODM), where you’re taking a product that already exists and putting your branding on it. For a brand new design product, it found looking online on sites like MFG.com is the way to go. In fact, I found about a dozen local (to Vietnam) manufacturers online and emailed them my design specification asking for a quote. I visited 4 (one in Saigon and three in Hanoi) and got quotes from all I visited. They were mostly around the same price, except one who was about a third cheaper. I went with the cheapest, because I’m still bootstrapping a proof-of-concept. There was some back and forth, which brings me to me second big lesson learned:
Pay for a design review with the manufacturer’s design team.
This is a huge time and cost saver, because things will come up in manufacturing, and having the manufacturer do a design review will prevent as much of that as possible. Of course, the manufacturer is going to look at your design before starting, but I think there’s a bit difference in actually paying them to do a detailed design review. They’re really going to put the time in and make some great suggestions, as I found.
I had my proof-of-concept milled, but for larger quantities you’re going to want injection molding. Molding can be done with plastic and metal, but doesn’t become cost-effective until at about 1,000 units. For instance, my part cost around $100 to mill for one or two units, but only at 500 units did the price per unit drop below $100 using a mold. At 1,000 units I am looking at around $80 per until. Of course prices will vary greatly depending on what you’re building. To give some idea of what I’m working with, my part is almost a kilogram of aluminum after milling, and takes about 4 hours of machine time, so pretty complex.
That’s it for today. The next step in this journey is to assemble a half dozen devices and get them in the hands of beta testers. Godspeed!