Self-driving: Right now, the biggest problem with self-driving car adoption is that when anyone dies at the hands of a self-driving car, there's a huge press outcry and the whole program gets shut down. That calculus changes dramatically if the other side is actively trying to kill you. Then, it becomes a huge advantage to not even send drivers into the war zone, and resupply with automated logistics lines. Instead of becoming an invention that potentially kills people, it becomes an invention that saves lots of people, because the baseline shifts from "nobody ever dies young" to "young people are actively trying to kill each other". It also gives a huge advantage to the side that adopts it, because they can resupply as quickly as they can build materials instead of losing costly human lives in the effort, and the vehicles themselves can often follow evasive routes that wouldn't be practical for a human crew.
Drones (like nuclear weapons before them) make war a lot more expensive. I guess only the top 5-7 armies in the world will have the technology and expertise to maintain drones. So a conflict like the one in Iraq will have a swift conclusion, but the calculus among the big nations will remain mostly unchanged IMHO.
Self-driving: I think the biggest problem with self-driving in this scenario is the very moniker "self-driving". But it is very likely there are vehicles being tested today that are programmed to avoid minefields, use sophisticated path planning algorithms to avoid detection, minimize fuel consumption etc. Sort of a jeep-sized roomba that traverses a region instead of your living room.
Self-driving: Can you elaborate on this a little more? You mean having self driving trucks on supply routes makes it much harder to break supply lines?
EVs: I think this one is already mainstream. Not in the sense that most people already have or are buying EVs but everyone from car manufacturers to bureaucrats are on the train already and it is only a matter of time.