Is anyone here currently building, or planning to build Stephen Hawes’ impressive Index PnP?
https://github.com/index-machines/index
I’d really like to build one, but I’m just waiting until the design settles down a little more. I think the project is getting really close. If you haven’t seen them, he has a series of entertaining videos documenting the design process:
https://www.youtube.com/playlist?list=PLIeJXmcg1baLBz3x0nCDqkYpKs2IWGHk4
I am interested in building a PnP machine. More for the challenge of it than anything else but also to incorporate any needed features into the next addition to my grblHAL breakout board line.
2 places where this sort of thing is discussed - you may already know these:
EEVBlog manufacturing sub forum
Open PnP group
I’m entirely on the same page - very interested if and when it stabilizes.
I’m in no hurry though, as I currently don’t have the space.
To pre-empt the discussion whether it makes sense: I’m fairly sure it doesn’t make economical sense to do so. But I love the idea of relative independence and looking at the collapse of quick, easy and cheap international services it is something I’m increasingly interested in.
I don’t need 0201 or the like on 8 layer boards - rather build a prototype on 0805 on 2 layer boards and iterate out errors quickly until I arrive at a final design. Then, maybe, I’d want to shrink the design for mass manufacturing.
I sure hope cheap/international assembly doesn’t go away. That would probably be due to a collapse of the global economy but we are living in unpredictable times.
I learned the hard way to always prototype with easier packages first. Reworking tight 402 and 603 placement is possible but truly not fun for me. There should probably be a thread on prototyping best practices.
For what it’s worth, I watched a few of Hawes videos. OK in small doses and a relatively low SNR. Not looking to be entertained, just informed. Guess I don’t need to see lots of hand soldering of parts.
I’m the same, very small runs (under 100) of boards with 0805 parts. Currently hand-placed because even the very cheap Chinese assembly options are too expensive. Although it would be fantastic to design something that was popular enough to have 1000+ assembled commercially 
I also wonder if there will be a market for selling very low volume assembly services? A bit like the early days of 3d printing, a local hobbyist with a PnP could offer 1-100 boards assembled cheaper than commercial services, and with quicker delivery? Not sure how it would work with components though. I guess it would have to be a mixture of holding stock of a small number of very common parts, and then supply of the rest by the customer?
I don’t know about offering services. I would never want that headache, because by definition the people you’re dealing with think differently than me. I am not going to fret over the odd misplacement here and there, someone who pays you definitely will.
So I’m budgeting time and resources only for private use. If I need income to pay for it, I just won’t do it.
If you want to run a business, don’t do it with a homebuilt machine. Even the low end Neodens and such are simply not good enough for production demands. Reliability is super important, every time you miss a deadline because of a broken PnP machine, your reputation takes a hit.