(just for info) How does Company design a low volume Handheld Device?

[Lord of nothing]

Hi
as the titel say how does Company build some low cost low volume Handheld Device who have internally some IO to connect Various thinks to the main System like a Barcode Scanner, Ethernet, DVB-T2, Wifi,...?
Are there some "ready to use" Systems from Company where the contain the main components on an PCB like a Mobile Phone this days would have? When I see how small the Phone PCB are I wonder if there are some "OEM Boards" to integrate them into there solution.
Thanks!

[petemate]

I'm not sure I understand your post correctly, but I guess you're asking about how a company is able to design a device with a specific function without having to "waste" time designing everything around that function..

The answer is, as usual, "it depends". Its quite easy(everything is relative) these days to design a system running some sort of embedded linux. If you don't need that level of complexity, its even easier to attach a small screen to an STM32 or other MCU to display some values captured from your specific sub-system. If you have the other engineering disciplines in-house anyway, you might as well design the whole device yourself. That being said, there are a lot of consultancies that will do this for you, so you can focus on your specific function. You basically tell them "i have this thing and i want that thing" and then they will design "that thing" for you, making sure "this thing" fits inside it.

For more complex system, you are basically building a normal computer and do the subsystem as e.g. PCI-express, USB or something, and talk to it through some drivers. Noone will design application-specific motherboards with CPUs, RAM, etc. anymore. Its off-the-shelf components that plug into the peripherals that are already present on the off-the-shelf components.

[Kilrah]

Raspberry Pi Compute Module? Or other SoMs (System on Module) more generally.

[Lord of nothing]

Like here: https://www.hytera.com/en/product-new/lte-broadband/poc-radio/pnc380.html
I dont think the produce a lot of them.
The same for the Payment Terminals from Chinese Sellers. 

EDIT: Or take any Bodycam from any Manufacture.

[jwet]

Most of my career I've designed low volume, scientific and industrial stuff and this was a constant issue.   You want to make a nice looking instrument but you don't have the volume to amortize a bunch of NRE (non recurring engineering- one time costs, tooling) over.  The other problem is that many of these specialty instruments have a long design life that make using an off the shelf solution difficult- the suppliers often come and go.  Here are a few strategies- it takes some creativity.  You also have to define how low a volume you're talking.  You can do a lot more at 1000 units/year than you can at 5 units per year.  The other thing to consider is to lower your expectations.  Its not realistic to expect a low volume gadget to be as polished as an Iphone.  Shoot for utility and quality.

1. There are modifiable industrial platforms that are designed for this purpose like the one you found.  There have been some platforms like the PoBox that have come and gone, or are now rare.  You'll have to find these, they will be from smallish niche companies.  If they are well thought out, you can take them, make a board for your special sauce, squeeze it into the package, write software and ship it.  The supplier needs to be stable enough for your product's life.  At very low, high margin businesses (sexy industrial), you can do a lifetime buy at intro and just hold them.  The down side is a surprise upside in demand.  There are some interesting stuff that is not exactly tailored for these uses.  The PicoCalc is an interesting platform that has a lot of function, looks good and could make the core of a custom instrument.  As you go along, do a make/buy analysis on whatever you're buying.  The engineering tools are so good today that you can make a lot of stuff.  The vendors like the Chip and uP Companies want to enable you.

2. There are package companies like PacTec and Bopla that make decent looking standard enclosures.  These aren't just lab boxes with aluminum covers from Radio Shack, they have gotten pretty stylish in some cases- look around.  These can be dressed up with  custom art work and custom keypads for volumes around 100 units cost effectively- getting keypads made and panel graphics is easy and pretty cheap.  At medium volumes (5K), these companies can put inserts into the  molds to customize rather than having to have all your mods done on a CNC- relatively expensive.  Extruded packaging is another good solution.  You buy a standard or custom extruded profile, cut to length and put ends on it.  You can make a custom extrusion die pretty cheap, (China), you'll have to take a minimum run of your shape.  There are companies that specialize in this stuff.  You can dress these up with custom ends and engraving.  Modern 3D printing can make ends that can look nice.   Workable for very small volumes and sturdy.   You then find one of more boards that will do your function and you make a custom board for your special stuff.  As in number 1, you have to find DIY type boards that are documented and open enough to write custom firmware.  Reverse engineering a closed box can be a big time.sink.  Things like writing a wireless protocols can easily take man year to recreate.  Look to uP vendors for libraries for advanced functions.  Be careful with licensing terms.

3. Attend some trade shows aimed at small manufacturers.  These is one in the US called "Design 2 Part" that gathers these specialty suppliers in one place.  Talk to them about your realistic volumes and see if they understand the problem.  There are companies that specialize in this stuff.  They have to make money, expect to pay them a living wage.  China has suppliers that you can find that will do things really cheap- like the PCB vendors?  I don't know these guys survive though I use them.  It doesn't seem sustainable.  Hire a creative Industrial or Mechanical designer with good CAD skills and some entrepreneurial spirit.  I've been lucky enough to find a few of these guys over the years.  They are generally very practical, creative, non degreed, jacks of all trades.  Many are comfortable in a machine shop and are proficient with modern CAD tools, Pro E or Solidworks.

4. If you have a applications where there  are a number of low volume related instruments, you can make an investment into a master platform that you could use on multiple projects with tweaks.  This can also give your product line some design continuity if done carefully.  You are essentially making number 1 above in house.  In some cases, you might be able to sell these as a product.  You have to be careful that you don't  equip your competitors but often in very specialized markets, its a don't care.

I hope this helps.  I enjoyed this kind of work, its challenging to make quality equipment at low volumes.

NB. Some Industrial Design houses like Whipsaw, Continuum and many others can work with you and have contacts with low volume and model makers that can make decent packaging.

[Lord of nothing]

Thanks for replay. 

I am wondering is it possible to 3D Print an Molde to see how Metal would form when there is a real one use to bend sheets into an form when use Alu Foil for Test purpose? 

[negativ3]

Justify going fully custom versus using available tech to do the job.
What is the cost of failure? Is it for in-house use only or for customer distro?
What happens if the custom widget designer quits or retires? Device lifecycle?

[jwet]

I've heard of this- 3d printing forming dies for sheet metal.  Poke around on line- I've seen YouTube guys doing it.

There is a danger in this stuff, you're an instrument maker, don't get too distracted making low cost low volume packaging.  If at all possible, move up the value chain and make specialized stuff that demands higher prrices.  You can't have low volume, commodity stuff and expect to survive- there is just too much stuff out there.