https://aseq.substack.com/p/inside-the-lucira-check-it-covid
They include an STM32, but the instrument is performing an Isothermal PCR molecular test which makes the inclusion of a microcontroller a requirement.
Obviously throwing out electronics like this seems quite wasteful. But overall our society is fairly wasteful when it comes to disposable electronics.
This might explain partly why they are all sold out :(
I’m just lucky we’re a non-electronics business that shipes electronics parts. If we were an electronics company, I would seriously consider a year hiatus, or switch to software contracting for awhile.
I do wish that there was an established recycling pipeline available for them though.
I bought 50 of these 2 weeks ago. Were 3-4 euro a piece.
There are some valid reasons for tests like this to exist. To quote later on in the linked Tweet thread [1]:
BTW it's easy to say this is wasteful, and it probably is, but just like with the pregnancy test, you should consider that there's three main problems that can affect the accuracy of lateral flow tests like this:
1. inconsistent lighting
2. incorrect timing
3. human error
having a 2$ computer chip and 50 cents worth of plastic & lenses removes all those sources of error.
And making it bluetooth removes a source of e-waste! it means it doesn't need to have a screen, it can just talk to your phone.
for example, consider that a bit more than 1 in 12 people have some kind of colorblindness.
Having a computer say "POSITIVE" or "NEGATIVE" is going to be easier to see then the uncertainty of "is that stripe red? can I just not see it?"
wait, so bluetooth is cheaper than paper?
A low-tech solution would be too boring or easy.
Seems unnecessary, but if people want to buy it…
Considering that this is an article used once for a few minutes the environmental cost for production and deposition is high and society shouldn't tolerate ...
Schools and food banks can hand out free self tests so there is definitely some kind of subsidizing going on here, but in a different way.
Germany has approved a much larger number of test kits for use, per https://antigentest.bfarm.de/ords/f?p=110:100:4793921863454:::::. It was >50 as of August 2021 according to https://marginalrevolution.com/marginalrevolution/2021/08/te..., including from American companies that the FDA has not approved.
I suspect that the supply-restriction effects of the approval situation go a long way toward explaining the pricing differences...
If you just want processing power, GigaDevices has an M4 at less than $1 last I checked. Unfortunately, it wouldn't work for the design we were working on as the standby power consumption wasn't great.
Seems like you could just about accomplish the same thing with a pair of NFC tokens, a photodiode, and some transistors. Basic components. Connect one of the two NFC chips to the coil depending on how much light is reflected off the test strip. When you scanned it with your phone the activated NFC chip would respond with a URL containing a unique ID which could be interpreted by the manufacturer.
They are cheap, last 10 years with a single battery, and do only what they need to do. Also secure. No crazy attack vectors. And easily symbolically verifiable. And I wrote tons of simulators and fuzzers for them.
Okay, I'm going to ask "Compared to whom?"
Nordic has been one of the better BLE chip manufacturers in terms of documentation, in my experience.
Most places use Keil tools directly from ARM. So, if programming these sucks, so does programming most embedded Cortex M4 chips. (No argument. Keil sucks. However, that doesn't mean that Nordic sucks any worse than anybody else).
And now you can program Nordic chips using Visual Studio Code.
All told I figured out how to make a reasonably complex chain of bluetooth devices work (largely by reading the source and comparing against old version API documentation to figure out differences), so it definitely could have been worse. But to imply that just because the documentation is BETTER than the competition that it is GOOD?
No. Its fine.
And you DO NOT have to use the Nordic SDK, there are lots of better and fully FOSS options — Apache Mynewt, libopencm3, Rust nrf-rs…
Also, it's not booting Linux at all. For these simplish one-purpose applications, you don't want the overhead and it probably would need more than the 24k or RAM onboard the nRF52810.
They could be using Zephyr though as it supports the Cortex M4 (it's a linux-like RTOS that can be heavily customised).
Embedded development is much different than what a lot of developers are used to.
The earliest machines used a light source and color filter to try and get the signal from the finished enzyme sandwich. Then lateral-flow came along and it was up to a human to make the observation. Now we're back to the computer doing it again.
https://www.cnbc.com/video/2021/02/01/us-231m-deal-ellume-at...
Maybe I'm Elizabeth Holmsing it, but I'm thinking it could have been done for 1/50th that price while still paying people well and providing them the resources they need.
Am I missing something?
I think this is a timeframe/certainty axis more than an efficacy/cost one.
Meanwhile I'm currently in Europe and as a consumer I can buy Antigen tests for around 3 Euro per test (would be less than 2 Euro if bought in bulk). As far as I know those prices are not subsidized.
I can only compare it to FreeScale chips, but one thing that pleasantly surprised me is the flexibility in pin-assignment. Instead of the pin-mux that we'd have to do on the MK24, I could just assign pins during usb/i2c init without any limitations. The SDK is quite elaborate, although sometimes a bit difficult to use in C++.
For most users, they prefer to have the chips cheaper and spend an extra few minutes of engineer time figuring out which pins need to be hooked up to what.
I can see for smaller volume devices where engineer time dominates, and devices that you want to be really field reconfigurable, the all-to-all mux is worth it though.
https://news.ycombinator.com/item?id=29637592 (Faking a Positive Covid Test (f-secure.com))
Meanwhile the paper versions work just fine.
It's not even different. Instead of having to interpret a couple lines you have software interpret the strip and tell you over bt on your phone. So much tech for so little.
https://aseq.substack.com/p/inside-the-lucira-check-it-covid
The best waste is the one that does not exist.
Why is there any kind of market for this?
You could do this with an analogue circuit and two LEDs.
Electronics must not be disposed with general waste, I'd imagine all Western municipalities have special recycling programs for electrical devices. Not sure if they'd be that happy to receive used covid test kits though.
We could subsidize research into biodegradables, though ...