4 ms·
> Or do you expect the people in the article to just write the drivers for their own implants themselves? No, but a community of volunteers might. ThreadX [0]
by awesomeMilou 2y ago
> Or do you expect the people in the article to just write the drivers for their own implants themselves?
No, but a community of volunteers might. ThreadX [0] is proof enough that an OSS community can keep a piece of software certified under strict security standards. I don't see why the same couldn't be achieved with software for medical devices.
[0] https://threadx.io/faq/#what-happens-with-the-existing-safety-and-security-certifications-for-threadx https://threadx.io/faq/#what-happens-with-the-existing-safet...
- HeyLaughingBoy 2y agoA couple reasons come to mind. (1) you can't separate the software from the rest of the device: it has to be Approved as a whole. (2) certification isn't enough. e.g., you could submit the software to a quality agency who would certify that it meets IEC62304, but then the entire device and its clinical test data still has to be submitted to the FDA, along with its Quality Record and all that (submission process, maintaining the Design History File, etc.) costs Real Money. And if you manage to get it approved for sale, then there are regulatory requirements around support and maintenance of the device that aren't going to happen for free. It's a nice idea, but unless you get funding from a philanthropic billionaire, I don't see it happening
- whythre 2y agoSounds like a good reason for these theoretical grassroots programmers to circumvent the regulatory agencies that are more of a hindrance than a help. ‘We cant do that because bureaucracy’ is such a non-starter.
- HeyLaughingBoy 2y agoIf you think that the FDA is more of a hindrance than an help, then I guess I have nothing else to say.