But cost of raw materials go up all the time. So the manufacturing cost goes up too, right ?
That is where standardisation comes in.
Right now Russia produces AESA radar modules for ground based radar for SAMs and for general scanning from portable small sets right up to building sized OTH radar for cruise and ballistic missile detection. They are starting to produce them for ships in radar antennas of all kinds, even vehicles like tanks will use them for APS systems to detect incoming rounds, and artillery can use it to track outgoing rounds to make sure they are going where they should, UAVs can use the to find targets and even as a datalink antenna to transmit data over significant distances, anti tank helos can use it to find targets, all sorts of missiles can use it for terminal homing, ships and even satellites can use it on an enormous scale.
Now each of these hundreds or perhaps thousands of different types of AESA modules can be custom made and totally different from each other in design, though likely made from similar materials in hundreds of different factories around Russia...
Or you could say... they all do basically the same job so lets standardise the design.
Russia is actually good at that... the submunitions in a Smerch 300mm artillery rocket are the same as the submunitions in a 152mm artillery shell, or a 500kg cluster bomb... they standardised the munitions for the job they do so they make enormous numbers and just fit them into whatever container they need for the weapon in question for the target required.
AESA radar antennas attach to systems... don't think of it as a "radar", think of it as a keyboard connecting to a computer network.
If you standardise the module design all the factories making them will be making the same thing... instead of needing 10,000 modules for ATGMs, and 20 million modules for 30 Armata MBTs and 500 million modules for the radar arrays for 3 new S-400 batteries... etc etc etc just make billions of modules and fit them as needed.
It means you can order materials and resources in bulk... it means all factories can pass any extra production examples to other projects if there is a production problem in one factory.
Over time the technology will improve, production speed will increase, prices will go down because of this... they might periodically update the design to make it better or smaller or cheaper and easier to make... that will effect all AESAs in production.
I had a photo that showed radar elements from the late 1990s and over a period of about 5 years there were four changes in design... the first picture showed an element that was a miracle... it was about the size of a pack of cigarettes and was revolutionary as it was a fraction of the size of the piece normally used, yet its performance was better. The other three pictures showed further improvements in design every year or so and the last picture was a chip sized element the size of your thumbnail that did the same job as the first component... but it was smaller, lighter, cheaper and actually did a better job.
The point is that they might not bother starting production of these components until they get to matchbox size for more critical uses... a building sized strategic radar system might not care even if it is desk top sized... it is needed... but refining the design and making it cigarette packet sized makes it even better.
Obviously sometimes different frequencies are needed... just look at the VHS and high frequency range AESAs of NEBO... it is simply not practical for them to both use the same AESA modules, but for all high frequency tracking radars it makes sense to unify the design... improvements in the design can then benefit all platforms using that module.
Once it gets small enough you can start putting it in things that would not normally carry radar, which gets interesting... imagine the squad machine gun with an under barrel AESA radar that would allow you to shoot targets through a dust or snow storm?
In fact the radar could be used to help guide rounds so they only hit the target and nothing else...
The point is by mass producing a standardised design you can buy materials in bulk, which normally reduces costs and interchangability greatly enhances logistics and support... look at motor cars... imagine if all cars were standardised so any door panel will fit any car for example...
Manufacturing is expensive if you don't produce very much.
The cost to build a factory is the same for making one thousand pens as it is for making 1 million... if you are making 4-5 pens then each one custom made would be cheaper than building a factory, but if you are making a lot, the more you make the cheaper it becomes as long as you know you are going to be making a lot.
Obviously if your factory is designed to make 500 pens a day then an order for 50 million pens is going to be a problem... if you knew you were going to be making that many pens then 500 pens per second might have been a better design goal.
If you design the factory with enormous production capacity but don't get the orders then you lose money... it is a question of getting the production capacity to match expected and actual demand... getting signed contracts is a must.
Of course if demand rapidly expands often it makes sense just to build another factory or more... with AESA radar modules having lots of factories makes sense too.
Making things like aircraft... or more particularly aircraft engines and submarines and large ships is very specialist work that needs both ability and money... money alone is not enough... Saudi Arabia will not start producing aircraft engines tomorrow even if they desperately wanted to.
Even the US will take half a decade from now to produce rocket engines as good as the ones they are currently buying from Russia and they have lots of pretend money to throw at it.
For a while Russia lacked decent UAVs and had very little experience.
They have spent a lot of money on their own designs and bought a few from countries with real experience with them, and have also gained their own operational experience with them... it cost money and time and effort to gain experience, but they are now seeing results, what 10 years after 2008 showed them how useful unmanned recon is.
The point I am making is that it is like fire... fire needs fuel, a source of oxygen, and heat or a spark...
Making a decent modern weapon needs money, expertise, and will.
You can have all the will in the world and you can even have the expertise, but without money not much happens.... Much like Russia in the 1990s.
You can have all the money in the world, and hire all the best thugs, but you plans to take over Syria and Yemen might still not work out despite really really wanting it.
And of course you could have the money but not the will to do it.