Link: https://en.defence-ua.com/industries/bundeswehr_wants_unmanned_60km_sph_based_on_tracked_rch_155_loras_to_fight_inside_kill_zone_but_cant_remote_fire_yet-20160.html

facepalm what the heck is the point of all the armor then? Sure make it unmanned capable but making it unmanned doesn't magically mean it doesn't need a crew it just means that now it is way more fiddly and tenuous whether the crew can properly control the vehicle when under jamming and wartime conditions than if they were in it...

...and now you have to armor and protect the location the crew are remotely operating the vehicle from anyways so?

what the heck is the point of all the armor then?

Same as before: To keep the vehicle operational under fire. Now it just has less meat sacks to protect.

[...] it is way more fiddly and tenuous [to] properly control the vehicle when under jamming and wartime conditions [...]

Are you under the impression that in modern SPGs there are still people cranking some gears to aim the device? It's all electronics anyways.

The task of artillery is to fire indirectly at predetermined targets. Compared to a FPV drone the only live information the system needs to do this is target coordinates. Unless there is 24/7 100% effective jamming going on, a brief window is all that is needed to transmit those. And no, jamming causing coordinates to be scrambled is not an issue. The problem of transmitting complete data in unstable environments has been solved ages ago. Every other parameter to create a target solution can be determined locally.

If there is heavy jamming a human crew can't operate the SPG from inside its armored body either: Artillery is indirect fire, without visual contact or communications there is nothing to shoot at. An unmanned/automated system even has an advantage in this case: Transmitting human language is highly inefficient compared to textbased information. So it can make better use of short windows of communication.

And as for reliability in wartime conditions: The whole premise of unmanned vehicles is that they can be abandoned if needed without having to spend time and effort to recover valuable crew. Yes, they are going to break down. So what? Send in the next one and mark the old one down for pickup at a later date when the area is cleared of hostiles.

...and now you have to armor and protect the location the crew are remotely operating the vehicle from anyways so?

Unmanned vehicles can be controlled at any given range depending on the circumstances: If there is no jamming going on, the operator can sit thousands of kilometers away at an office desk or meters below ground in a bunker.

If there is no communications, as I wrote previously the whole concept of artillery is breaking down anyway.

If there is no jamming going on

So the place you are planning to applied repeated pressure with 155mm artillery shells the size of your torso you are expecting NOT to have saturated enemy jamming? What the hell?

The article handwaves this away with saying fiber optics cables could just be used but then you have spent all of this time uparmoring an automated artillery system worth millions of dollars to make tactically defeating it simply a matter of snipping a thin cable, a pair of scissors from your kitchen drawer could do it....

To be clear I think building an automated capability is entirely reasonable and further I think it is reasonable to design all large vehicles as optionally manned (so long as they are not allowed on public roads which are far too chaotic) going into the future but please understand how this way of framing the problem is critically missing the point of having highly trained highly motivated human beings respomsible for machines that can change the shape of entire wars near singlehandedly under the right conditions.

I am not sure you got my point: If comms are jammed a human crew has to fiddle their thumbs as well, because they have no way of knowing where to shoot either.

And if communications are available an automated system can be faster and you don't risk valuable human assets.

but please understand how this way of framing the problem is critically missing the point of having highly trained highly motivated human beings respomsible for machines that can change the shape of entire wars near singlehandedly under the right conditions.

I can't help but imagine this being a comment by a kamikaze pilot being distraught by getting replaced with a drone.

I am not sure you got my point: If comms are jammed a human crew has to fiddle their thumbs as well, because they have no way of knowing where to shoot either.

Yes they do, the crew know the context they are entering into a fire mission on, they know the friendly units in the area and how they tend to behave, they understand how the enemy has been behaving in that area lately, while the crew may frequently be given fire missions with little context for operational security reasons that doesn't mean the crew isn't highly aware of what of the tactical and strategic role they are playing in the combined arms system.

You are terribly terribly wrong by assuming removing the human is a direct improvement, you can't really remove the human so the only pertinent question is how fragile, convulted, fiddly and breakable is the chain of command between the human crew and the vehicle itself?

If the human crew is nearby or in the vehicle, they can much more easily establish contact with a friendly unit relaying them a fire mission in the area and make it happen quickly, no third party is needed, if all hell breaks loose so long as that howitzer crew remains dynamic and in contact with friendly units in the region (which under heavy jamming may not be through intermediary communication systems like satellites) they will know what to do without being needed to be micromanaged by command like an unmanned vehicle would.

A human crew, at the point of breaching in the frontlines, whether on the offense or defense will make decisive independent decisions as best they can when it comes down to it if they have to and know there is no other choice but for them to take responsibility and make a choice for everyone. Robots? Well they kind of can, if you think automation is trustable ok but automation does the worst with edge cases and new information/patterns... which is going to by definition be the experience of your defenses that becomes most existentially important during the critical period of an overwhelming attack on your defenses....

Exhibit A https://lemmy.ca/post/72646855

https://united24media.com/war-in-ukraine/operation-vivaldi-sent-ukrainian-armor-deep-behind-russian-lines-straight-at-drone-operators-23617

As a minor point, TCP works badly if there is more than a small amount of packet loss. It suffers from nondeterministic delays as well as head of line blocking.

There was a lot of research around better transport for specific tasks in the 1990s and 2000s (and probably long before and after that), for Voice over IP as well as for video streaming and the like.

Your point that there are well-studied approaches for lossy networks is valid! Just not TCP. 😄