Morning.
The F-35 may have passed a phase of testing but still has areas of vulnerability. Firstly such a Heath Robinson VSTOL reconfiguration has to be highly vulnerable to battle damage.
Secondly, as part of a continuously active logistics network
using ALIS, it's highly vulnerable to cyber attack. It may be possible to safeguard the network (though working in the business, I doubt it, it's a constantly changing battlefield), the problem is can you prove it?
Thirdly, there is the complexity of the actual
software/firmware on the aircraft itself. The same issues arise, both of V&V of the system and being able to prove that such a complex system is safe both from bugs and cyber attack?
What does it really matter if the VTOL system is 'susceptible' to battle damage any aircraft that takes hits in those areas will be in a lot of trouble and how do we define susceptible I can't imagine an F18 being hit amid ships is very pretty either. If the aircraft is badly damaged enough that it can't land vertically and is still flyable then it would be diverted to somewhere where a conventional landing is possible. Or the pilot would have to ditch the aircraft same as any other aircraft.
The greatest chance of loss would be in the landing and take off phase which is where a lot of the harrier accidents and loses seem to have occured with the caveat that the F35B is under computer control on a vertical landing not the manual control method of the harrier with it's wide margin for human error, reaction time, spatial awareness and ability to multi task 100%
every time.
As far has hacking the ALIS network goes the obvious thing to reduce the ease of access would be to use a customised encrypted messaging protocol with rotating keys. Then maybe use radio datalinks for nodes that are completely seperate from the web and then utilising military satellite bandwidth to get data to the regional maintenance facilities and then on to the 'mothership'. Ultimately it seems to me that if you limit access to the system through using a seperate network to ao any other traffic and/or make that access reliant on a physical connection on say a military base with lots of big hefty men with guns and systems to purge machines when being over run then you'd get a long way to lessening your problem.
I'm not being funny but any large complex distributed software engineering project has these exact same issues with VV&T from the signalling system of a rail network to the ground control station of a satellite and everything in between. Thats why these things take years and years and years and why the software drops are so slow. But that's why the discipline is called software engineering and not 'programming'. Do you try to mitigate your risk and progress or never attempt anything 'complex' and stagnate?