Probably from the perspective of the people on either end, too, over time - I cannot imagine that the RoE and confidence bounds were designed without considering the expected costs of blue-on-blue incidents compared to the alternatives.

I mean, it kind of has to fail-deadly to be of any use at all - since what's a tank with no IFF?

"Firing solution ready. Execute?" seems like an unambiguous query. But I know nothing about this stuff. And I wonder what happens if the operator doesn't respond. I'm guessing that nothing happens. Because who wants a tank with an unconscious or dead operator shooting at everything?

The enemy doesn't - It could still be useful for the own party.

Are they remote-controllable?

It'd be great to get some insight into this.