Overall, more transparency by the armed forces and industry is needed to scope effective governance and utilisation

Ideology is a key component in this heterogeneous network of actors, ranging from ingenuity and curiosity to an expression of support with an actor in an armed conflict (e.g. Ukraine or Myanmar), over libertarian beliefs about the right to bear arms, to violent extremist ones

Diverse non-state actors are designing, manufacturing and supplying self-made armament, for which they often use AM (3D printing), such as (parts of) firearms and drones, stabilisation fins of grenades, bomblet shells, etc.

Embracing AM as a military capability means that repairability must be enabled by design, both technically, i.e. procured technology is designed so MRO can be independent of the OEM, and legally, i.e. purchasing contracts consider a right to repair job of the armed forces, wherever possible

Within several militaries, AM is being tested (e.g. in Ukraine) and in use for MRO to respond to supply chain issues. While mobile solutions already produce "part of consequence" for armed forces, those projects are often tests and don't equal a scoped adoption of the tech into the organisation

The main challenge for serial adoption is production efficiency in light of high costs, broad certification requirements and testing regimes, non-destructive quality control and lack of material diversity, etc.

Across military domains, trials of additively manufactured items are documented, but seldom their output. Combined with vague and sparse information by manufacturers, AM's utility for primary production is challenging to verify; however, AM has become significant for signalling technological edge

"This Is Not What I Signed up for": Sociotechnical Imaginaries, Expectations, and Disillusionment in a Dutch Military Innovation Hub" by Sofie van der Maarel, Désirée Verweij, Eric-Hans Kramer & Tine Molendijk

And no word about costs attached to maintenance, recovery, training, or problems in regard to the reliability of the computer vision or other ML-based functions in a non-permissive environment etc.

Anyways, I wrote about the limits autonomy in the land domain. For more nuance, and less rant see 3/3

or the challenges in designing multi purpose robotic systems. Further, the idea of cheap and expendable robots, that can fulfill the expectations (autonomous, resilient, multi-purpose) in the land domain is far from reality and both engineers and the armed forces are quite aware of this 2/3