I'll buy your electronics to feed our robot (dayworkx.com)

by skholinn 41 comments 103 points
Read article View on HN

41 comments

[−] skholinn 50d ago
We’re building a robotics system that recovers reusable electronic components from retired hardware. To run experiments we’re looking to buy electronics that are *retired or obsolete but still contain valuable components*.

Examples:

* servers, networking gear, routers, switches * laptops / workstations * telecom / industrial / embedded boards * lab equipment electronics * obsolete or end-of-life hardware with populated PCBs

We’re especially interested in hardware that *is no longer useful as a system but still has valuable chips or components on the boards*.

Typical pricing depends on the hardware, but we often pay *$20–$200+ per unit* for things like servers, networking gear, or laptops depending on what’s inside. Happy to buy bulk lots.

Based in the Bay Area; we can arrange pickup locally or pallet shipping within the US.

If you run ITAD, recycling, refurbishment, repair, or have retired hardware sitting around, email:

sava@dayworkx.com

Even rough descriptions like “a pallet of old switches” are helpful.

[−] girvo 47d ago
This is very very cool :)

And is _way_ better than when I'm forced to do this by hand, I'll say that much haha

[−] Joel_Mckay 47d ago
There are e-waste companies with inert hot-air centrifuges. Essentially, they extract the RoHS solder, clean it, and resell re-certified product back to manufacturers.

The chips simply drop off into a bin, and no ethical company will sell customers used silicon. There were a few folks that ended up getting a few years in prison for that con (small China groups would sand and laser mark old chips with modern lot codes), and hence why many US recycling plants shred the chips for recovery of precious metals.

It is a serious safety problem, and having personally been stung by locally sold counterfeit/used stuff with BS compliance documentation... it sometimes means months of lost project time figuring out what happened. Always direct sample from the manufacturers whenever possible. =3

[−] femto 47d ago
I can't see a problem, as long as the chips are not fraudulently resold. Beyond not using a resource in the first place, reuse is the gold standard in sustainability.

As an engineer, I wouldn't use second hand components for prototyping. When prototyping you need to eliminate as much uncertainty as possible. I'd consider using second hand components in production, provided there is a cost advantage, supply is reliable and my production line includes a test that would pick up faulty components. Even then, I'd be monitoring failure rates and reverting to new components if elevated failure rates caused costs. There's an argument that (well handled) second hand components might even have a lower failure rate than new as they have been burned in.

I'm guessing this company is targeting specialsied repair rather than production. Sometimes complex parts are no longer manufactured and the only option is second hand (often at a premium price).

[−] Joel_Mckay 47d ago

>I can't see a problem, as long as the chips are not fraudulently resold.

In general, most components are only rated for 2 to 4 re-flow heating cycles before internal damage occurs. On some components the initial re-flow cycle brings the component into the rated tolerance, and for others the PCB forms a bimorph cantilever that physically fatigues the chip contacts/leads.

Production yields are only part of the Infant Mortality Phase of the bathtub curve.

Some components do get more stable with age if and only if left alone, but you can count those on one hand if you still have all your fingers. That is also a 3 hour pedantic conversation no one wants to have.

I am secretly a sentient turnip... =3

[−] femto 47d ago
That's a fair point: that heating due to repeated (de)soldering can cause degradation.
[−] skholinn 47d ago
We want to get the data on that. The more boards we process the better we know the failure rates. Do you have an intuition of what exactly degrades?
[−] femto 46d ago
My intuition in this area is based on chips having a specification on maximum soldering temperature and duration. I'm not sure to what extent that is cumulative. I gather the vulnerability is the bonding of the gold whisker wires to the pads on the silicon, but you would want to check that.

Apart from the absolute temperature, chips have a recommended heating/cooling cycle, including heating/cooling rates. That suggests that differential expansion is a factor, which would likely be cumulative (more cycles = more likelihood of fatigue and damage).

The above is intuition, not the hard data you want.

I think what you are doing is a great idea (effectively demanufacturing). I'm hoping you can solve the practicalities, which as far as I can see are quality assurance and being able to guarantee a steady supply of components and a price point below new.

Any plans to retape the components so they can be put though a pick-and-place machine, or are you looking more towards manual rework? I can see that there is room for innovation in efficient ways to get components off boards at volume, as most component removal is in the form of manual rework.

[−] phyzome 47d ago
...OK, I'll bite, what's the "sentient turnip" bit about?
[−] skholinn 47d ago
the way we see it is that with robotics and coding agents we can offer much comprehensive and traceable tests. anyone can send any hardware they have, we help understand what is reusable, how to verify it and ship the parts back.

curious—what specific failure modes or uncertainties would you want eliminated before you'd consider using recovered parts, even just in production?

[−] HeyLaughingBoy 46d ago

> no ethical company will sell customers used silicon

Since when? My first EE job was in 1988 and back then "pulls" -- components pulled from circuit boards were an easy way to get cheap parts. We'd get mailers every week from various companies listing what they had available. The electronics recycling industry was/is huge.

[−] skholinn 47d ago
exactly!! we aim to have comprehensive and transparent protocols of testing after the extraction which is exactly what is hard without general purpose robotics. if you have any pointers on how to make that happed or any pain points from your experience please send me a note to sava@dayworkx.com !
[−] skholinn 47d ago
thanks! how often do you have to do it by hand and how long does it take it?
[−] userbinator 47d ago
China has been doing this sort of recycling for literally decades, at a massive scale.
[−] jollymonATX 47d ago
So I send in massive 60 drive jbod pcbs and you pay me more than 55c/lb? That's current clean pcb rate at any recycler. Boards are ~8lbs ea. Usually just tossing them unstripped to a muncher that pays 35c/lb for the whole 55lb jbod works out way better for time labor.
[−] imtringued 46d ago
I guess the prime target for this would be USB-C controllers? Ubiquitous and expensive enough to justify building a machine and yet versatile enough that you could find a second hand market for them.
[−] jarbus 47d ago
This is really cool. In the demo, how do they just yank the chip off the board? I'd have assumed it would be somehow soldered on or something.
[−] AndreasKromann 50d ago
Awesome concept !
[−] skholinn 47d ago
thank you for all the support so far! We've got a lot of inbound and started processing it. there are a lot of interesting questions on this thread and i'll address them soon.
[−] nikhilyadala 50d ago
Is it only in the USA?