PLA Next Generation Main Battle Tank

Alkyone said:

I still have a hard time imagining how automation can replace the workload of a 3rd crewman. Of course, the driver's role should be focused on driving. And certainly, you can replace the loader with a mechanical loader. That leaves the commander with... literally everything else, the most important roles being commanding the tank, aiming the gun and spotting enemies.

With a bit of imagination, I can see how automation can spot enemies via a series of sophisticated sensors, picking out infantry and tanks and showing them on a videogame-like HUD. With all these clips showing how advanced surveillance AI has become in China, I can absolutely see this system working out. Still, ideally, you want a commander that's on the lookout for the whole battlefield situation, rather than tunnel vision himself in a gunner role.

There's also the communication aspect. Now I've never operated in a tank, but I served as a signal op, and many of our folks also trained as gunners on IFVs, where they fulfilled both the role of gunner and radio operator. I would imagine a similar separation of tasks also exists onboard MBT. Comms doesn't feel like something you can leave to automation either - unless the driver take over that role since he's now seemingly sitting next to the commander?

Also, a lot of people here also mentions that 2 crew is the minimum to operate a tank, and one should have an extra crew on board if not just to keep the machine operable in an emergency - actually, I can see this MBT being operable by even just 1 crew, not too dissimilar to the Strv 103, if they duplicate controls for both crew members.

But still, as others stated 2 man crew just feels lacking even in the modern battlefield. I look at western prototypes for their 4th gen, where their 4th crewman now either operates an autogun turret (useful against helicopters or drones possibly) or controls a drone of some sort, and it feels like there's definitively potential there.

Click to expand...

But no one thinks jet fighters need one more crew member, whether it is for air superiority or acting as a bomb truck? All the attack helos also only have two crew members and no one asks for adding a third one. Even strategic bombers are having less crew members.

So what if future MBTs are just like how jet fighters or attack helos work except on ground?

All the situation awareness problems can be solved with better sensors and fusion of data from friendly units. In battle fields, such tank units will always be supported by drones and/or manned airborne sensors. Reliability will have to be much better so tanks will require much less maintenance and often can be offloaded to crews at home. Endurance will also need to increase a lot to cut down frequency of replenishment.

Then you only need two crew members to operate an MBT, a driver and a gunner. And they can switch roles to cover each other when needed. From there, one can even start to dream about single-operator MBTs.

IIRC, these are what 毛明 thinks. He is the chief designer of Type 99A. He shared his thoughts on future MBTs in an interview by CCTV. That video has been posted in this forum, likely in the Type 99 MBT thread or this thread.

The text report of the interview can still be found:

War may start high tech, but if a short high intensity conflict ends up prolonged the logistics and systemic support might degrade overtime and crew might find themselves fighting a war that start to resemble the early 90's and in that case 2 man crew might become disadvantaged.

Maybe have a platform that only need 2 crews to operate but reserves spaces for 3 members. Have layout where if the 3rd crew is not needed then maybe it can be used as space for more ammo or fuel etc.

I think 2 man standard crewing is possible in wide and flat terrain with AI and added on maintenance robots to make up the loss of the 3rd crew in terms of information load, maintenance and situational awareness, while the tanks instead work in pairs with every second tank having the 3rd crew member being the commander that can analyze datalinked information from the UAV and second tank, and make strategic decisions.

Units determined to be operating in terrain e.g. forests and mountains where tanks are more easily isolated or ambushed will have the 3rd member restored.

Sure the space used to maintain the 3rd crew will still be there, but the casualties will be reduced overall. I am also skeptical of how much more space you can save by removing the 3rd member considering that you are already reducing the existing space by making the turret unmanned.

On topic of "2 crews bad", I am not disagreeing. With the constraint of current gen it makes sense. 3 crews are needed for situation awareness using current gen technology. But we are talking about something like 2030. Imagine if each tank is followed by 5 drones each feeding videos to crew's helmet. AI auto identifies possible targets 360 degree, gunner just need to confirm. But we can't all rely on AI. It appears 2 people are operating a tank, but in reality dozens of drone operators are working behind line marking targets.

TK3600 said:

Zzzzz Soviet Union is 80s. This is 40 years later.

Click to expand...

40 years later than 1980 doesnt change the fundamentals of armoured warfare. AI also still haven't matured to the point where it can replace important crew. You need AGI for that which is still a paper-only idea.

Can't comprehend e-commerce and drone warfare today.

Click to expand...

Leningrad under GABTU designed Object 299 to operate with drone attack helicopters, LOS-comm UGVs, mass launched NLOS swarming missiles. Your believe on 2-man tanks sounds more like what US generals would masturbate themselves to during FCS rather than reality. So please stop with the barebone insults and do actual research into Soviet cyberwarfare and radioelectronic combat theories.

this is not my idea, this is what China is working on

Click to expand...

Yes, and I'm saying that anyone in the PLA that is actually serious at the idea of 2-man tank is utterly stupid and detached from reality. Given that the PLA hadn't done tank warfare for the last 4 decades and has no experience at operating tanks in high-intensity combat of any kind I'd let this pass. Still amusing though.

I trust these people with PhD knows better and there are reasons to try. There could be more engineer vehicles.

Click to expand...

PhD is meaningless without experience. TACOM built laser APS, mini railguns for tanks, 1kg PGMs and liquid propellant guns before. They built unmanned tanks, unmanned M113-based STUGs, unmanned combat engineer vehicles and deminers, see-through fused panoramic sight and AI-controlled loitering munition swarms. The same guys also felt for FCS. Given that China's largest achievement in building MBTs sofar is rehashing foreign techs into VT-4A1 I seriously doubt their ability to solve the problems associated with 2-man tanks.

They could use live data link for situation awareness

Click to expand...

Datalinks only transmit... data. You still need someone to see the information that has been processed and displayed, then make a decision. That last part, the AI still can't do, and this issue will remain for a long time.

There could be other soldiers outside tanks help maintnence

Click to expand...

If you are seriously proposing support forces to dismount and fix your vehicle in heavy combat, youre delusional. Let's say you are in combat against an enemy tank company, and your track is hit, or suddenly your autoloader electric drive is knocked out of place, or fragments bashed your engine grill and smokes are stuck. Meanwhile your mates are getting blown up, one has their turret removed, and needs to retreat. Who will the support guys help? Manpower is limited. You could just retreat to defilade and have 2 guys out to roll back the tracks, or pull out stuck shrapnel, etc.

See, I'm not against R&D of any kind. If the PLA wants to know what will their combat effiency be if they shifted their ORBAT to a logistic-dominant forces and reduced frontline manpower, comparing to their existing doctrine, be as they wish. What I'm against is actually bringing 2-man tanks into mainstream service.

TK3600 said:

As for why not on the hull, NERA was found to be less effective vs long rod penetrator vs KE than air gaps + steel.

Click to expand...

Anything that's not plain steel or spaced armour or ERA is NERA. The T-72B's bulging plate arrays are NERA, as referred to by the owner of the Tankograd blog.

In other words western tanks using NERA on the hull as opposed to ERA panels is not 'better protection'

Click to expand...

Yes it is?
NERA is intergral armour. External ERA only provides protection for 2-3 times max with the heavy ones like Relikt or FY-4.

Abrams's lower hull was notoriously weak vs KE until the recent upgrades

Click to expand...

Warthunder again. Abrams hull was equally as protected as contemporary ( Leopard 2A4) hull. What the early Abrams' array was strong at was defeating HEAT rounds. It was 600mm in LOS thickness alone. That vulnerability stemmed from underestimating Soviet APFSDS. Later arrays provide magnitudes more KE protection.

bulging plate vs reflecting plate

Click to expand...

Both are the same. The "reflecting plate" is used to refer to the T-72B's NERA array. "Bulging plates" is a generic description of NERA, since practically all NERA bulges, like the Leopard 2A4's or T-90A's.

SanWenYu said:

But no one thinks jet fighters need one more crew member, whether it is for air superiority or acting as a bomb truck? All the attack helos also only have two crew members and no one asks for adding a third one. Even strategic bombers are having less crew members.

Click to expand...

5th gen fighters are a completely different beast from MBTs. This is apples to orange.
Furthermore.
NGAD will not be the classical fighter anymore. It will fight as a sensor platform and accompanied by AI-directed UCAVs and AWACS. So alot of the work is offload from the fighter platform. The US Army is proposing Apache's successor under FVL to be base on the V-280 but armed with ALEs, EW pods and NLOS ATGMs and manned with 5-6 crew.

SanWenYu said:

So what if future MBTs are just like how jet fighters or attack helos work except on ground?

Click to expand...

So a MBT with 30mm of titanium armour with a masted radar, 30mm cannon and external missile pods? Apples to oranges.

SanWenYu said:

All the situation awareness problems can be solved with better sensors and fusion of data from friendly units. In battle fields, such tank units will always be supported by drones and/or manned airborne sensors.

Click to expand...

Assuming the perfect scenario- the downfall of any military planning.
What if your OPFOR jam local comms and you cant talk to your friendly forces?
What if your OPFOR deploy obscurant and the sensors can't see anything?
What if your OPFOR deploy SPAA and shoot down your UAVs?
Are you willing to accept that if your plan goes the other direction, just for a tiny bit, your tankers will be blind, deaf, ineffective, and dead within hours?

SanWenYu said:

Reliability will have to be much better so tanks will require much less maintenance and often can be offloaded to crews at home.

Click to expand...

Reliability scales inversely to capability. What is more reliable, VHF or flags? Digital thermals or bulletproof glass optics with etched reticle? SATCOM or maps? Exposed radar panels or hard armour? And then, what is more capable?

SanWenYu said:

Endurance will also need to increase a lot to cut down frequency of replenishment

Click to expand...

If you cant guarantee logistical support, you are dead, or at least sort of. You dont have logistics? You dont have fuel, food for your tankers, ammo, some electronics parts to repair a faulty drive or circut board, or APS/smoke loads, or clothes, or practically most other essential items that will get bashed pretty quickly because you stored them in your bustle racks, and those racks got hit by airburst 152mm.
Stowage is relative to size. The more you can store, the larger your vehicle is. Now wait for that drones that didn't see a T-72 anywhere but can see your Maus standing out like a white elephant. So, instead of making your tank a glorified hamster, improve your logistics, and guarantee that you will have a supply line towards your frontline troops at all time.

SanWenYu said:

Then you only need two crew members to operate an MBT, a driver and a gunner. And they can switch roles to cover each other when needed. From there, one can even start to dream about single-operator MBTs.

Click to expand...

Safezone driving can be substituted by an AI autopilot. Combat manuevering is much more difficult. TTB tested interchangable crew station and turns out it's really darn hard. But doable.
However the role of the MBT is far more complex than just shooting and driving. And most of those roles are really stressful.
I know people will tell me " so why didn't the US or Russia try 5-man tanks". It's not needed, and what's under question is the feasibility of having tanks manned by only a crew of 2. I shall quote an answer on a Quora question on the Renault FT:

The driver? He was in charge of:

• driving.

The tank commander? He was in charge of:

• leading the driver;
• looking out to find targets while leading the driver;
• acquiring targets;
• aiming the gun;
• shooting the gun;
• reloading the gun;
• communicating with others tanks with flags.

And it was somewhat great for him to don’t have a radio, since he would have been the radio operator too.

Click to expand...

You can remove point 1, 2, 4, 6 and 7. Unfortunately you would also have to add these:

1. Operate UAVs and UGVs.
2. Direct fire missions.
3. Communicate and coordinate infantry.
4. Same with tanks.
5. Same with IFVs.
6. SIGINT.
7. Jamming.

It's pretty hard. OMT had tankers trying out 2-man, 3-man and 4-man configuration. The result is that they can fight effectively with 3 man, can do it better with 4 but 2 is only capable of basic tasks and is hated the most.

yeetmyboi said:

40 years later than 1980 doesnt change the fundamentals of armoured warfare. AI also still haven't matured to the point where it can replace important crew. You need AGI for that which is still a paper-only idea.

Leningrad under GABTU designed Object 299 to operate with drone attack helicopters, LOS-comm UGVs, mass launched NLOS swarming missiles. Your believe on 2-man tanks sounds more like what US generals would masturbate themselves to during FCS rather than reality. So please stop with the barebone insults and do actual research into Soviet cyberwarfare and radioelectronic combat theories.

Yes, and I'm saying that anyone in the PLA that is actually serious at the idea of 2-man tank is utterly stupid and detached from reality. Given that the PLA hadn't done tank warfare for the last 4 decades and has no experience at operating tanks in high-intensity combat of any kind I'd let this pass. Still amusing though.

PhD is meaningless without experience. TACOM built laser APS, mini railguns for tanks, 1kg PGMs and liquid propellant guns before. They built unmanned tanks, unmanned M113-based STUGs, unmanned combat engineer vehicles and deminers, see-through fused panoramic sight and AI-controlled loitering munition swarms. The same guys also felt for FCS. Given that China's largest achievement in building MBTs sofar is rehashing foreign techs into VT-4A1 I seriously doubt their ability to solve the problems associated with 2-man tanks.

Datalinks only transmit... data. You still need someone to see the information that has been processed and displayed, then make a decision. That last part, the AI still can't do, and this issue will remain for a long time.

If you are seriously proposing support forces to dismount and fix your vehicle in heavy combat, youre delusional. Let's say you are in combat against an enemy tank company, and your track is hit, or suddenly your autoloader electric drive is knocked out of place, or fragments bashed your engine grill and smokes are stuck. Meanwhile your mates are getting blown up, one has their turret removed, and needs to retreat. Who will the support guys help? Manpower is limited. You could just retreat to defilade and have 2 guys out to roll back the tracks, or pull out stuck shrapnel, etc.

See, I'm not against R&D of any kind. If the PLA wants to know what will their combat effiency be if they shifted their ORBAT to a logistic-dominant forces and reduced frontline manpower, comparing to their existing doctrine, be as they wish. What I'm against is actually bringing 2-man tanks into mainstream service.

Click to expand...

On topic of Object 299 I knew it was a sort of Armata predecessor. Never knew it had sensor fusion and drones integrated. Would love read more on that.

When your tracks are blown up, engine cooked, autoloader jammed, you bail like what Russians are doing right now. No amount of crews is fixing that while under fiire by enemy tanks.

Leopard 2 in Ukraine got tracks blown by mine. Surely 4 crews can fix a tank under fire? No, they could not. It was subsequently destroyed by artillery.

Maybe it is possible if you are US doing low intensity warfare you can take your time and repair on battlefield. Not in any peer combat.

yeetmyboi said:

40 years later than 1980 doesnt change the fundamentals of armoured warfare. AI also still haven't matured to the point where it can replace important crew. You need AGI for that which is still a paper-only idea.

Click to expand...

No, you really don't need AGI replace crew. Current day AI can do way more than you think, and we're also not talking about only LLM AI.

Although, yes, today in 2023, I would say that AI would not be 'sufficent/good enough' to actually have a tank go from 3 to 2 people, especially not if we're gonna add more 'stuff' for the tank to do.

But what about AI in 2030? 2040?
The next gen tank for the PLAGF is gonna come around out~2030s and will be used into the 2040s likely early 2050s.

yeetmyboi said:

Yes, and I'm saying that anyone in the PLA that is actually serious at the idea of 2-man tank is utterly stupid and detached from reality. Given that the PLA hadn't done tank warfare for the last 4 decades and has no experience at operating tanks in high-intensity combat of any kind I'd let this pass. Still amusing though.

Click to expand...

LMAO, this sentence completely makes YOU to be the stupid one and uninformed.

"The PLA hasn't done tank warfare in 4 decades, so they know nothing about what tanks would face in a high-intensity combat situation" my ass lol.
They know way, way more than you, in fact, even way more knowledge about it than you would get even if you were to join US military and become a tanker or the likes, because quite simply, they are a big organizations withs lots of people (research, strategists, actual commanders, tankers etc.) that has it as their full time job to research and improve their own tank warfare.

They have tanks, anti-tank weapons, mines, drones etc. as well as troops, units, brigades etc.
With all of that, they can do realistic exercises and in fact, I think it's very likely that learnt pretty darn little from this current Ukraine war, because they already knew most of the lessons to be learned.

Not to mention, they can get intel from various sources, including, directly from the Russian ground forces lol.

yeetmyboi said:

Datalinks only transmit... data. You still need someone to see the information that has been processed and displayed, then make a decision. That last part, the AI still can't do, and this issue will remain for a long time.

Click to expand...

Lol, AI can already 'kind of do it'.
Just go and try feed GPT4 a dataset, and it can point out data as well as do a 'data analyse'. Is it perfect or the likes? No.
But is it still useable and can be very good? Yes.

In other words, it is definitely possible to have AI, even current day, do some analyse or process data, even if it might be very large and scattered.
And from doing that, it can give suggestions to a tank commander or the likes.

yeetmyboi said:

See, I'm not against R&D of any kind. If the PLA wants to know what will their combat effiency be if they shifted their ORBAT to a logistic-dominant forces and reduced frontline manpower, comparing to their existing doctrine, be as they wish. What I'm against is actually bringing 2-man tanks into mainstream service.

Click to expand...

The PLA will have multiple tank proposals/ideas to choose from, and it's very likely that they will actually get their hands on testing both a 2-man version as well as a 3-man version (likely near or around 2030).

If they conclude they rather like having a 3-man tank? Then that's what they are gonna get.
But if they do think the 2-man tank could work (with the 3rd man not removed, but placed behind or w/e), then it's something that they very well might get.

Oh, and these decisions will involve lots of people, testing, exercises etc.
It will be a 'highly scientic process', involving lots of people, up and down (generals, scientists/designers, strategists, actual tankers etc.)

Michaelsinodef said:

"The PLA hasn't done tank warfare in 4 decades, so they know nothing about what tanks would face in a high-intensity combat situation" my ass lol.
They know way, way more than you, in fact, even way more knowledge about it than you would get even if you were to join US military and become a tanker or the likes, because quite simply, they are a big organizations withs lots of people (research, strategists, actual commanders, tankers etc.) that has it as their full time job to research and improve their own tank warfare.

Click to expand...

Please actually quote my words instead of twisting it around. I stated "has no experience at operating tanks in high-intensity combat" which is a whole world of difference to "know nothing about what tanks would face in a high-intensity combat". If you think generals reading powerpoint reports can amount to any sort of tangible improvement in existing doctrine I'd suppose you tell that to Russia who had 8 years to prepare.
It's reality that the PLA still hasnt operated MBTs in combat. It's literally just that. They trained under controlled environment. The Soviets had 30 years to train their troops, and they still got squashed in Afghanistan, because instead of NATO artillery dumping DPICMs which would be countered by barrages of TBMs they faced a determined resistance force employing assymetrical warfare tactics. And also learned thay their logistics are shit, and logistics dictate war.
Look, I'm not disparaging the PLA in any way. I'm just stating facts, facts that had significant historical value. The PLA has no, zero, nil or whatever synonym you can find in 100+ languages on actually conducting centralized armour warfare. The only nation to have ever done so is the US in 1991, and they lost whatever knowledge they obtained then after 20 years of buying SOF toys. The PLAGF modernization has improved their artillery, operational fires, combined arms coordinations, battle engineering et al to world class, they are theoretically better than practically every other army bar America's BUT they, recognizable, haven't put a major focus on improving their MBTs and their employment of said assets.

If you disagree with what I said. Please, point out the following:

• The PLAGF has improve individual MBT sensing, endurance and decision making ability in complex urban terrains under bombardment/EW and platoon coordination/manuevering to the level of US tankers in 1991/ second Fallujah.
• The PLAGF has improve vehicle engineering ability to retrieve and repair MBTs during combat with pace.
• The PLAGF has demonstrated the ability to secure supply routes, provide reliable SHORAD cover, rapid removal of complex minefields/obstacles and deliver appropriate logistical support within schedule against stealth standoff PGMs.
• The PLAGF has demonstrated the ability to correctly employ armour forces to thrust, defend, scatter and escort while mantaining acceptable attritional ratio.
• The PLAGF MBT forces have demonstrated the ability to create/emply complex terrains and engineering to provide tactical advantages in battle.

Modern combat is far more complex than " I read today's sitrep dayum" and suddenly realizing which are you are deficient in and then suddenly improving that area. It take experience, literally, to do that.

Michaelsinodef said:

They have tanks, anti-tank weapons, mines, drones etc. as well as troops, units, brigades etc.
With all of that, they can do realistic exercises and in fact

Click to expand...

You know what is realistic exercise?
USAF officials turning off jammers because they were literally killing radio stations thousands of kilometres away.
No exercise are remotely close to reality. No exercise could and have simulated troops morales, faulty comms, airburst TBMs, constant artillery barrages. trench warfare, Soviet-style DIY, lack of force concentration, fog of war, low supply, wrong range extrapolation or just plain simple jammed rifles because the enemy missed your chest but hit your gun instead.
Yet Ukraine is infested with all of these problems.

Michaelsinodef said:

Just go and try feed GPT4 a dataset, and it can point out data as well as do a 'data analyse'. Is it perfect or the likes? No.
But is it still useable and can be very good? Yes.

In other words, it is definitely possible to have AI, even current day, do some analyse or process data, even if it might be very large and scattered.
And from doing that, it can give suggestions to a tank commander or the likes.

Click to expand...

You literally just restated what I said in that quoted section lmao.