A bit of disclaimer: I'm a Ukrainian army officer and I've spend this year planning and coordinating artillery fires. No kind of crack specialist - just a mobilized signals reservist actually - yet I saw here a lot of stats and cases and it changed significantly what I see as plausible future ground combat. I'll put it here in case somebody find it interesting and, if I'll survive, it may be a starting point to tinker with Aurora further.
The main thing - this year was a warfare revolution in full swing, drones became major (not auxiliary) combat tool. Before, I wasn't completely sure it's coming the way it seems, now it's a fait accomply. There may be another and less predictable revolution in a near or more distant future, yet returning to the past is, well, very unlikely.
Artillery was a major damage dealer throughout nearly all the industrial era. Now it's over. The guidance electronics became so capable, compact and cheap, that it's just no more need neither to saturate the field of battle with indirect fires just in hope that some shell will find some valuable target in the area nor to spend time and shells with adjustment fires to catch a regular target there. You spot a target and it's not covered - you just send drones to kill. You spot a target and it's covered - you send a heavier drone to either blast the cover or mine the exit. It's faster, more reliable and cost less comparing to spending 5 to 50 shells. It's also less risky - howitzer fires are very easy to spot, while drone starting points and their crew dugouts are not.
There were concerns that strike drones are vulnerable to EW. Well, they are. Still more effective then any other spotting and kinetic tool in most cases, even in EW-saturated area, because since your signals in this area are suppressed - it's usually not only drones, you also cannot shout to your infantry, and so you have nobody to spot for your fire support at all. The only solution is more EW-resistant drones - with proper FHSS modules, directed repeaters, towed optic cables, etc. Modern brigade-level recon drones - both Ukrainian domestic and European ones - are resistant enough, the opponent usually cannot suppress them even when they are 15-25 km deep over the opponent's lines, and it's just proper FHSS modules, nothing more. I know less about the russian drones, the general impression is those are mostly less capable, yet while our EW units are severely underequipped and vulnerable to enemy missiles - it's not enough of the tech gap to give us significant advantage. The US-designed drones are... strange, mostly useless and overpriced as hell, yet I think it's just like it was with their tanks at the start of WWII: it's not that it's very hard do it proper, it's just no perceived need to do it until Pearl Harbour, so, well, they'll fight having outdated drones will they fight somebody serious at all.
Another and more serious limitation is weather. Here are two factors. Bad visibility is actually just irrelevant - you cannot spot through heavy fog or rainfall/snowfall, drones or no drones, and without spotting neither strike drones nor arty cannot hit their targets without enormous ammo expenditure. The relevant weather factor is wind - infantry and armoured and traditional Air Force are shrugging it out, while Mavic-like drones cannot. Yet it's like muddy weather for tanks: the early models were extremely vulnerable, the makeshift ones still are, yet since you start producing proper tanks - they are mud-able enough and definitely more able to run through mud then any other advancing elements. The same with drones: start producing properly designed military-grade ones and it has to be a tornado to deny their operations, and their strike cost is still 5 to 25 times less then a needed barrel ammo to hit nearly any target.
The most serious limitation is currently actually dense vegetation and buildings. The thing is, the cause of drone efficiency is their "leanness" - no heavy shell, no redundancy, not much durability, small cheap motors. That's why they are cheap enough to spend literally in millions, sweeping out infantry like hell. Though, bump a branch or a wall with it - and it's down. There are some obvious ways to fix it - ring-covered props, better materials, etc. - yet I'll expect 15 to 30 years of R&D before you can send a drone through thickets like it's a forest hawk and hunt down somebody hiding there. Buildings are actually simpler - the Boston Dynamics androids are already capable enough to open doors, and climb stairs, move through the rubble, etc., while holding weapons, yet androids are just a new type of infantry. Flying strike drones are definitely not.
The major advantage of infantry was always it's low cost. It's no more. Light strike drones became much cheaper then infantrymen. Try to field an effective stormtrooper - and their harsh long training and weapons and equipment cost is around the same cost as a pair of drones and their operator with their kit, still capable of killing this trooper quite quickly and reliably anywhere except thick bushes and a dugout he'll be hiding uselessly, and dugout and bushes are no guarantee too. Without costly training and equipment troopers are a mob, you cannot make a breakthrough with them, nor use them as an occupational force, garrisons or something alike. So, it's forever, it's not a temporary doctrinal failure you can overcome just by procuring some R&D and making new regulations. Actually, I'll bet further R&D will just deepen the gap, making strike drones relatively more effective in any habitable terrain. Infantry are for garrisons only - to sit in bunkers and dugouts and basements under ruins, and never ever advance without overwhelming tech and fire superiority, end even then prefer to send drones.
It's not that barrel artillery is dead. Unguided MLRS are - not enough precision even to plow large positions, too much propellant expenditure, extremely easy to spot during fire, too much volume to armour and so easy to destroy with nearly anything. Traditional mortars are doomed too - aiming too slowly to hit mobile targets any kind of reliably, mine flight time too long, still too easy to spot from drones during fire. Automation may help with reaction time, yet flight time is still bad. Heavy mortars may find some niche with a mix of guided mines and cheap bombardment irons, yet it's troubled, range and flight time are still bad and the heavier mortar the harder it to dug in and hide. Yet long-barrel artillery has their stable niche, because it's kinetic strike has a capacity to be really fast. Strike drones are cheaper, yet much slower in delivering the warhead - they are mostly deep subsonic, cannot be supersonic without just becoming shells or missiles. Artillery shell is deeply supersonic from the start, fly time 2 times less then mortar mine, 10 times less then strike drone. So, you spot an important moving target and cannot hit it - you call artillery. Yet not the traditional one - this one needs preps (with the crew sprinting out of their dugout) and adjustment fire (at least 3 shot-correct-shot iterations), so no less then a minute to hit the target with low chances to surprise it. Strike drones are not much slower comparing to this, even the slow makeshift ones. Yet if it's a properly automated arty fire with guided shells - it'll reliably hit the spotted target from the first shot, and it's just 15 to 60 seconds ToT and a good chance for lethal surprise. Yet it's not a bombardment anymore, it's more like superheavy sniping. Bombardments are to destroy large fortified and urban areas only, and that's where long-barrel arty is also capable, even the traditional towed one with full crews - good propellant efficiency, long range, shots are too fast and chip and durable to stop them with any kind of AA fires. You need to cover your arty with robust EW and AA, and dug them all in before starting bombardment, because bombardment needs time, and artillery is easy to spot during fire, and EW and most of AA are easy to locate while they are active too, so the opponent'll try to massacre them with strike drones. That's why you'd hate to have short-barreled arty to bombard anything close to peer opponent: less range, more flash, so much harder to weather counterbattery fires and strike drones.
Tanks are not dead, howsoever troubled. You still need something to move through fire and wire and mines, and since it's not infantry, then it's tanks. They are to spot inside the battlefield smoke and dust, kill anything on the surface and move further, break the defense line, and do it under unending drone and missile strikes. You cannot do it with unarmored vehicles - they'll be massacred with strike drones even easier than your infantry. You cannot do it with weeled armored vehicles - they are still too easy to stop with light strike drones or mines and finish off while immobile. So what you do need is tracked, all-around-spaced-armoured tracked vehicles with mine plows. No other vehicle type will survive advancing through the modern battlefield long enough to be effective. It's very tempting to keep or even enlarge the main tank gun, yet this temptation is dangerous. Long-barreled heavy gun with it's shells and propellant charges are not only too heavy, but also hinder armoring and survivability. I'm sure it's worth rearming tanks with autocannons only - much smaller turrets, fast enough to track and shot down inbounds, still able to blind enemy tanks instead of trying to kill them all alone.
Traditional APCs are troubled alot, because traditional mobile infantry is dying. Heavy armoured or lean and fast - APCs for deploying heavy assault drones, powerful receivers and so on - well, yea, we need it, yet it's a question if it's more effective to do with APCs or heavy copter drones like our Vampires (aka Baba Yaga) or even something heavier, yet still flying and unmanned. And carrying your infantry to storm enemy positions - the further the more it looks like sacrificing your men and combat vehicles for nearly nothing. To advance is actually to move your receivers towards enemy positions, then to kill all the surface opposing elements, and then to clear the dugouts and thickets, while your forward elements are moving your receivers and clearing the surface further. Of these tasks we (currently, in Ukraine, as an observable fact) still need infantry only to clear dugouts and thickets (and so secure logistics for further advance), and it's only because we have not enough shells to destroy dugouts and clean thickets, and I hope in 15 to 20 years nobody (except religious fanatics) will need troopers for the latter too. People are too slow, weak and fragile for the modern battlefield, and no training can change it.
Attack helicopters are dead. It's an expensive tool of giving direct fire weapons longer fire line by just lifting them up. It's obsolete as a concept, because indirect fire guided ammo became plenty and capable, so no more need to just lift a platform up - you use either cheaper and stealthier surface or unmanned platform or faster and longer ranged winged one. Drones capable of running down helicopters dosens of kms over the frontline became usual, which means even medevac copters are also troubled - they need either more speed or defensive turrets and still be troubled just because you cannot make helicopter stealthy for thermals, and drones with thermals became ubiquitous.
Traditional combat aircraft definitely is more then just alive. It gives you kinetic start - you can toss relatively cheap guided bombs and longer ranged fast missiles instead of burning a lot of hard propellants to build the same kinetics by launching missiles from surface. It gives enormous operational flexibility - you can concentrate or disperse your strikes, choosing which exact sector of the frontline you'd like to penetrate, and it's in hours, not days or weeks, it's very hard to predict where the opponent's Air Force going to strike next hours and so where exactly to deploy your defenses and what to hide urgently. Cheap drones cannot intercept jets and missiles - you need to arm them with proper missiles to do it, and it's not cheap anymore. You need to defend the air bases, yet it's more then possible - we do it for years even having no effective interceptors and no strategic depth (all our territory is under strikes every day), and our airforce is still operating, and we have now more Soviet-designed bombers then there were at the start of full-scale war, which looks like miracle and is not - it's hard and costly, yet possible and worth doing, Air Force kicks like hell if you know how to do it properly.
