4 ms·

Related: https://spectrum.ieee.org/bionic-eye-obsolete https://spectrum.ieee.org/bionic-eye-obsolete > Their Bionic Eyes Are Now Obsolete And Unsupported It's

Related: https://spectrum.ieee.org/bionic-eye-obsolete https://spectrum.ieee.org/bionic-eye-obsolete > Their Bionic Eyes Are Now Obsolete And Unsupported It's for this reason that I believe all medical implant firmware and required associated software should be FOSS. Nobody should ever lose the use of something embedded in their body because a company goes out of business.

telesilla 4y ago
I can't see a requirement to be FOSS working due to those who fund such companies expect to have dividends, even universities. But at least if it's published, and with stronger right to repair laws, there is some chance rather than none.
- ckemere 4y ago
  You could imagine filing firmware updates with the FDA to be released if the company ever stops providing support?
- walkerbrown 4y ago
  Perhaps mandate a trigger? In the event of bankruptcy, implanted medical device firmware and hardware designs enter the public domain. Patients’ interests should take precedence over creditors and investors.
alerighi 4y ago
Yes, the code of these devices should be 100% FOSS. Also the argument about defending intellectual property doesn't make sense: the software will probably do trivial things and be very simple, that is what you are selling is the hardware itself, and that is already covered by patents. So making the code public will allow people to find a replacement in the eventuality that the manufacturer stops supporting the product. This should be imposed by the governments, in my opinion, also for another reason: when the company that makes the hardware device stops supporting it, in countries with a public healthcare system it's the state itself that has to support the patient. How can the healthcare system provide assistance to something that doesn't know internally how it works? How can it go to another company to ask to fix the device?
- 9wzYQbTYsAIc 4y ago
  In the case of the above implant, open hardware specifications would have been needed: the stock of replacement hardware (the computing device worn on the body) is limited. The Spectrum article gives an example of a patient that had to work with his doctor and hunt for unused devices to replace a device that was damaged from a drop.
- taneq 4y ago
  > the software will probably do trivial things and be very simple, that is what you are selling is the hardware itself, and that is already covered by patents. How do you figure? In the case of something like an insulin pump or a pacemaker, sure, but once we start interfacing with nerves, especially once we start trying to communicate with them properly instead of just blunt-force stimulating a bunch of them, these devices are essentially software. That's not to say I necessarily disagree with your conclusion. For devices that depend on patented hardware, opening up the software still works because the hardware effectively acts as a copy protection dongle. For other devices, or for cases where the software contains critical 'secret sauce', what we need might be something more like the automotive right-to-repair legislation. Even if the device is closed source, the manufacturer must provide appropriate documentation and tooling to allow third parties maintain the device for its intended service life (which, of course, is hopefully a bit longer for a brain implant than for a hatchback!)
  - alerighi 4y ago
    FOSS software doesn't mean that it has to be free. We have to remember this. It's perfectly fine to ask for a payment to have the source code. Also the fact that the software is free it doesn't mean that it can't still be covered with patents, for example a lot of patented software is free software (think about video codecs, for example, there are a ton of FOSS implementation, still it's patented). Basically you patent the algorithm, not the implementation, that is anyone that uses it still has to pay you royalties, but anyone can produce it. If you think about it, it's the same thing with drug formulas: after a period of exclusive then any company can produce it, of course paying the royalties to whoever owns the patent. It can be the same for software. Of course we have to talk about algorithms that are not trivial, as it's the case on most medical devices (most firmware just does very trivial things, that could be in principle be done by dedicated analog circuitry)

telesilla 4y ago

I can't see a requirement to be FOSS working due to those who fund such companies expect to have dividends, even universities. But at least if it's published, and with stronger right to repair laws, there is some chance rather than none.

ckemere 4y ago
You could imagine filing firmware updates with the FDA to be released if the company ever stops providing support?
walkerbrown 4y ago
Perhaps mandate a trigger? In the event of bankruptcy, implanted medical device firmware and hardware designs enter the public domain. Patients’ interests should take precedence over creditors and investors.

alerighi 4y ago

Yes, the code of these devices should be 100% FOSS. Also the argument about defending intellectual property doesn't make sense: the software will probably do trivial things and be very simple, that is what you are selling is the hardware itself, and that is already covered by patents. So making the code public will allow people to find a replacement in the eventuality that the manufacturer stops supporting the product. This should be imposed by the governments, in my opinion, also for another reason: when the company that makes the hardware device stops supporting it, in countries with a public healthcare system it's the state itself that has to support the patient. How can the healthcare system provide assistance to something that doesn't know internally how it works? How can it go to another company to ask to fix the device?

9wzYQbTYsAIc 4y ago
In the case of the above implant, open hardware specifications would have been needed: the stock of replacement hardware (the computing device worn on the body) is limited. The Spectrum article gives an example of a patient that had to work with his doctor and hunt for unused devices to replace a device that was damaged from a drop.
taneq 4y ago
> the software will probably do trivial things and be very simple, that is what you are selling is the hardware itself, and that is already covered by patents. How do you figure? In the case of something like an insulin pump or a pacemaker, sure, but once we start interfacing with nerves, especially once we start trying to communicate with them properly instead of just blunt-force stimulating a bunch of them, these devices are essentially software. That's not to say I necessarily disagree with your conclusion. For devices that depend on patented hardware, opening up the software still works because the hardware effectively acts as a copy protection dongle. For other devices, or for cases where the software contains critical 'secret sauce', what we need might be something more like the automotive right-to-repair legislation. Even if the device is closed source, the manufacturer must provide appropriate documentation and tooling to allow third parties maintain the device for its intended service life (which, of course, is hopefully a bit longer for a brain implant than for a hatchback!)
- alerighi 4y ago
  FOSS software doesn't mean that it has to be free. We have to remember this. It's perfectly fine to ask for a payment to have the source code. Also the fact that the software is free it doesn't mean that it can't still be covered with patents, for example a lot of patented software is free software (think about video codecs, for example, there are a ton of FOSS implementation, still it's patented). Basically you patent the algorithm, not the implementation, that is anyone that uses it still has to pay you royalties, but anyone can produce it. If you think about it, it's the same thing with drug formulas: after a period of exclusive then any company can produce it, of course paying the royalties to whoever owns the patent. It can be the same for software. Of course we have to talk about algorithms that are not trivial, as it's the case on most medical devices (most firmware just does very trivial things, that could be in principle be done by dedicated analog circuitry)