It’s a common refrain from new clients: “we hired a design firm to do a previous project. They made a nice design but when we found out what it was going to cost to build we had to scrap it.” There are several techniques to keeping a new design’s costs in check, but one of the most useful is the concept Bill of Materials or concept BOM.
Engineering teams since the dawn of the industrial revolution have used bills of material to list the parts of an assembly often listing the part number, description, quantity and other useful information. (Fellow geeks please check out this cool resource for old maritime drawings.) The problem is that most teams don’t use them until near the end of the design phase when they start building prototypes or putting together documentation for production.
I prefer to get the design team thinking about the BOM right at the concept level, before we even know what specific parts will be involved. You will never control costs unless you have a cost target, so you’d better have that in mind when working on concepts. Even during the concept phase, your team can list functional subassemblies or functional areas. Take for example something like a handheld drone controller. You might not know of how many parts the chassis will be made, and you might not know if there will be one or two PCBs in the final product. But, you can figure there will be a housing subassembly, electronics assembly etc. So your concept BOM for a handheld controller might start something like this:
|Housing Subassembly||$ 25||Injection Molded|
|Electronics Subassembly||$ 120||2 PCB’s|
|Connectors||$ 120||3 x MIL-SPEC|
|Joysticks||$ 100||2 hall effect|
|Display||$ 250||5″ LCD w/touch|
Then you’ve got your cost target. I’m usually thinking in terms of cost to manufacture the parts not including labor at this point instead of selling price. Think about what you can realistically allocate to each subassembly and of course what you think they might actually cost. Now here is where there’s really no replacement for experience. You will need to take a swag at costs for the subassemblies and you’ve got to keep annual volumes, likely manufacturing methods etc in mind. I’ve got tooling costs in the back of my mind at this point. If you’re working on low to medium volume projects, you’d better be talking with the client about rough expectations for non-recurring expenses for tooling and fixtures. And seriously, annual volumes can take a part’s cost up more than 10x when you get into low volume manufacturing so don’t take that lightly.
As you progress, your BOM might start to look like this:
|Housing, Top||$ 10||Inj Mold, EMI coated||$20k|
|Housing, Bottom||$ 10||Inj Mold, EMI coated||$20k|
|Door, Battery||$ 5||Inj Mold||$8k|
|Main PCB||$ 40|
|Radio PCB||$ 80|
|Connector, Power||$ 35||Glenair|
|Connector, RF||$ 25||Fischer|
|Connector, Video Out||$ 25||HDMI, sealed|
|Joystick, Left||$ 80||Otto Hall Effect|
|Joystick, Right||$ 80||Otto Hall Effect|
|Button, Power||$ 5||Sealed, non-illuminated|
|Display, LCD, 5″||$ 225||800×600 w/touch panel|
|Hardware||$ 10||Stainless with black coating|
Now sit back and use all the experience available to you to check if your assumptions are reasonable. Early in my career I made it a point to seek out the experienced engineers where I worked to ask for advice. When you’re just starting out nothing says “I respect your opinion and experience” to a veteran like honestly asking for help. (Thanks Dave, Gordy, Bill, George, Matt and Dennis). Another technique is to get two or three team members together and each write down your individual cost estimates and see where they vary. If you’re off by much there might be missing expectations, or one of you just has more experience in that area.
If you’re really stuck with a new material or process you’re not familiar with, you can use a couple of other tricks. Your best bet is to contact a manufacturer who makes the parts you’re looking for. If it’s a custom screw and you’ve never done anything but bought standard hardware, contact a screw maker. They might be able to help if you ask for a rough-order-of-magnitude (ROM) quote. Most people don’t like to give a price over the phone, but if you describe your need, you can usually say something like: “So I need to know if this is a part that’s gonna cost ten cents, a dollar or ten bucks?” You’ll usually get a reply that gets you close enough for the concept BOM.
Another less accurate trick is to find a product that is made with a similar manufacturing technique and in similar quantities. You can figure the manufacturing cost to be somewhere between 3-6x lower than the retail price unless it’s a commodity product then the markup might be lower.
Yes, we are relying on a lot of experience here. But, even with loads of experience, your new design can diverge from the cost target drastically if you don’t start thinking about it early. Once your concept starts to develop, the BOM is a living document and you’ll start splitting up your subassemblies into real parts and you can give those swag estimates too and keep yourself in check.
Lastly, once the product goes into manufacturing, make sure to close the feedback loop. I recently had a client complain about the cost of a part we specified. Having a concept BOM I was able to look at the $12 he had for this rubber bumper and the $1.50 I had figured and I immediately knew we had a sourcing problem. The client got their cost from a retail supplier, not directly from a manufacturer so I was able to point them in the right direction and help them find a better source. Sometimes though the feedback will be ugly. You need to hear the bad news too so you’ll get better each time around.
Work on a concept BOM at the beginning of each project and update it throughout your design process and you will have better luck hitting your cost targets.