There are many assumptions here that should be questioned.
Source code reviews are highly imperfect ways to ensure stable and accurate software, and good ones are extremely hard on the developers involved. Techniques like test driven development and paired programming offer a much better solution at lower cost.
New medical devices are released all the time and they have new code operating them, even if that general type of device has been in use for decades. New models with new or modified code have new bugs.
Perhaps owners of electronic devices that have caught fire or misbehaved in other physical ways have learned to start inquiring about manufacturing, mean time between failure and other manufacturing and quality issues.
I have worked in the medical software industry for thirty years as a developer, and was at one time an employee of Medtronic. I have a Medtronic pacemaker/defibrillator embedded in my chest which can be remotely accessed and controlled. I am professionally qualified to study and understand my device's software, development and testing methodology, and security issues - but Medtronic declined to share with me their source code when asked. The technical manuals for my devices which appear to provide all necessary information for hacking my pacemaker/defibrillator are available online.
I think that more can and should be done with oversight of medical device manufacturers and their software than the FDA currently requires, but this is true of all mission critical software like military and aerospace systems as well. The problem is neither uppity lawyers nor uncaring medical device manufacturers but instead the way we build software. Anyone with personal experience in the software industry who relies on a programmable medical device but who is not concerned over the accuracy and stability of the software running it is not thinking clearly.
do a barrel roll. fsu football fsu football do a barrelroll bérénice marlohe bérénice marlohe google offers
কোন মন্তব্য নেই:
একটি মন্তব্য পোস্ট করুন