0
\$\begingroup\$

Could somebody tell me what the rough steps would look like? I think I have a good idea. I have put things into practice, and it's working and so on, but it's very ugly.

I guess, first I would have to build a prototype which looks much more like the final product and is not as ugly and so on. But what are the steps exactly, how do I best build a prototype and then move to mass production?

PS: 'Mass production' might be a bit misleading here: I just want to have 500 pieces at first. I just have no idea where to start and how the process looks like.

\$\endgroup\$
2

1 Answer 1

Reset to default
4
\$\begingroup\$

Just a list of the main steps I went through:

  • Product concept.
  • Features selection and limitation.
  • Component selection, mainly CPU and big peripherals.
  • Schematic entry prototype.
  • PCB design prototype.
  • Writing, testing & debug real-time embedded code.
  • Writing, testing & debug development GUI.
  • Assembly prototype.
  • Re-design schematic final product.
  • PCB design final product.
  • Assembly prototypes final products.
  • Updating, testing & debug real-time embedded code.
  • Updating, testing & debug development GUI.
  • Selecting Chinese manufacturer.
  • Interfacing with Chinese manufacturer.
  • Preparing manufacturing files (e.g. Pick and place coordinates).
  • Schematic entry Test-jig.
  • PCB design Test-jig.
  • Writing, testing & debug Test-jig self test code.
  • Writing self-test failure analysis manual.
  • Writing, testing & debug Manufacturing self-test GUI.
  • Arranging for EU EMC conformity testing.
  • Arranging for EU ROHS certificate.
  • Negotiating for EU WEEE directive.
  • Applying for Registered Trademark.
  • Arranging EU/UK import registration.
  • Arranging packaging.
  • Mechanical outfitting Test-jig.
  • Writing manual.
  • Writing, testing & debug Drivers for C.
  • Writing, testing & debug Drivers for Python.
  • Making You-tube instructions & advertise movies.
  • Registering and and setting up webpage
  • Adding & testing last minute extra features.
  • Programming and shipping first 1250 products.

And then you have to set aside plenty of time for customer questions, updates, complaints, returns etc. for the next four years.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ hm, depending on what you build, you might need to add patent research, negotiation, avoidance to that list; or, even, writing, applying for and paying for a patent. \$\endgroup\$
    – Marcus Müller
    Commented Jan 16, 2018 at 22:10
  • \$\begingroup\$ Also, I'm curious: What is the product you developed? \$\endgroup\$
    – Marcus Müller
    Commented Jan 16, 2018 at 22:11
  • 1
    \$\begingroup\$ Not in my case. I may be clever but not that clever. Biggest worry was that the product would be just successful enough that the far east market would copy it before I had my return of investment and all my money an effort would be lost. (Can't tell you what it was. Would be advertising. Still ongoing, I am now in the end phase of the sales curve.) \$\endgroup\$
    – Oldfart
    Commented Jan 16, 2018 at 22:14
  • 1
    \$\begingroup\$ (well, having worked with a few people trying to bring their patents to buyers/licensees: you don't have to be very smart to be holding a patent; neither does the patent have to be very smart) \$\endgroup\$
    – Marcus Müller
    Commented Jan 16, 2018 at 22:15
  • 1
    \$\begingroup\$ if you plan to buy RPis, add hardware and sell the result make sure you can actually buy the model in volume (pi-zeros for example don't seem to be buyable in 100s or 1000s) \$\endgroup\$
    – Taniwha
    Commented Jan 16, 2018 at 22:29

Not the answer you're looking for? Browse other questions tagged or ask your own question.