

The main defense a rail network has is that it is fairly easy to repair track at an industrial scale so long as you can clear the area and rebuild, and as a result that even though a train is a very very vulnerable large loud target, it is difficult to know WHEN to ambush a train because you might just be standing in the middle of nowhere for hours and the train never comes before hostile patrols make contact with you.
Under this logistics system however rail networks become constant sources of intelligence on enemy movements, this seems like a disastrous idea to me given the sophistication and skill of Ukranian UAV and unmanned ground vehicle operators. Also, it wouldn’t take much to stop a whole kilometer long length train of these unmanned logistics carts, you just need to blow the one up in front with an FPV drone and the rest of them are stuck. This is the war equivalent to placing a traffic cone in front of a self driving car in order to immobilze it, and by virtue of not using a truck or a train with a human driving it Russia leaves itself open to massive amounts of logistical disruption this way… which is what loses wars ultimately.
As you point out Russian troops are screwed if they aren’t near train tracks but that also means Ukrainian intelligence can assume the Russian troops are having to tactically come within a certain distance of tracks to resupply which makes their movements massively more predictable and easier to disrupt. You just look at a satellite image and start drawing lines to the closest rail lines in enemy territory and extrapolate from there…
Trying to predict where an MRAP or APC will rush in much needed supplies to a heavily suppressed defensive unit on the otherhand is much harder to do both from the increased mobility standpoint but also from the standpoint of needing to muster a far greater degree of precise firepower needed to knock out the logistics vehicle even if you can predict where it will be.
Yes, but also consider the basic wisdom of it, I believe the origins of the Super Tucano extend back to WW2 fighter craft like the P-51 mustang being evolved into trainer aircraft occasionally used for ground attack. Everyone gets excited about the bleeding edge of unmanned aviation development but from the perspective of the Super Tucano as an airframe… I mean… sure you can develop a fancy new heavy lift unmanned ground attack and sensors platform… or you can just take a Super Tucano and modify it.
The same thing is happening with the Lakota helicopter sort of being phased out as a manned aircraft in direct frontline military use in the US military while the same time?? the same exact model helicopter is being heavily considered for an unmanned heavy lift vehicle (well medium/light utility helicopter but from the perspective of unmanned systems “heavy”) by the US military. It is enough to make your head spin like a helicopter rotor.
I think the important thing to remember is that the media, general public, military and techpress will want to obsess over the legitimately terrifying new developments in unmanned aerial (also ground and littoral) vehicles because they are novel and intimidating. From a practical perspective however, the best platforms for unmanned navigation systems are manned aerial vehicles with a long and developed history of maintenance, documented behavior in emergency maneuvers, and extended third party market for modifying the same platform for a variety of purposes… while in the field.
To put it simply, once you get past a certain size of horse, it doesn’t really matter if a human is riding on the back of it or not, what matters is what stuff the horse is doing, whether that stuff is a stupid idea or not and whether it is being supported properly to do said horse stuff. So yeah, Super Tucanos, Aero Sharks, Lakota Helicopters… all of these types of platforms are immediately the best candidates for the future of manned or unmanned aerial surveillance, reconnaisance and defense at depth.