August 7, 2026
In January 2023, a GPS-guided 155mm round supplied to Ukraine was recording confirmed successful strikes about 55% of the time. By August, that figure had fallen to 6%.
The weapon had not changed. The battlefield had. A Ukrainian assessment of nearly 3,000 rounds, reported by the New York Times and the Washington Post, traced the collapse to Russian electronic warfare as jamming refined faster than the munition could respond. The same reporting describes a satellite-guided glide bomb losing accuracy within weeks of its introduction, with observed misses ranging from about 19 metres to close to a kilometre. Eight months. That is how long it took for a precision weapon that worked to become a precision weapon that mostly did not.
Now set that against the other clock. In July 2026 the US Government Accountability Office found that across the Department of Defense's 104 costliest weapon programmes, the average time to deliver a capability has risen to more than twelve years. Twelve years to field. Eight months to counter. That gap is the defining problem in defense today and no amount of additional spending closes it. The question is no longer whether a weapon works on the day it enters service. It is whether it can keep working on every day after.
The lesson from Ukraine is adaptation velocity
The most cited lessons of the war in Ukraine are the return of mass artillery and the rise of the drone. Both are important. But the more fundamental lesson is how quickly the battlefield now evolves. Throughout the conflict, both sides have continuously rewritten software, swapped sensors, changed frequencies, altered guidance logic and rebuilt tactics in response to what the other side did last month. Advantage has rarely belonged to whoever fielded the better system first. It has belonged to whoever improved theirs fastest.
The GAO reached the same conclusion from the opposite direction. In testimony to Congress, it found that leading companies "use iterative cycles to design, validate, and deliver complex products with speed," activities overlapping under continuous user engagement and testing while in a linear acquisition, cost, schedule and performance are fixed early against requirements set years in advance. Even on the rapid pathway Congress created, most programmes could not show how they would deliver a fieldable minimum viable product inside its five-year window. Ukraine is running that experiment under fire. The acquisition system is still debating it.
Iteration is an architectural decision, not a policy
Most weapons cannot be iterated quickly because they were never designed to be. A guidance update triggers lengthy requalification. New electronics require redesigning the surrounding architecture. Production lines built for frozen configurations resist change once they are running.
The result is predictable. Threats evolve. Weapons do not. Designing for iteration means making different decisions from the very beginning. It means modular subsystems with clearly defined interfaces. Open architectures. Digital engineering. Software-defined functionality. Continuous validation. Production processes that assume change rather than resist it. None of these decisions can be bolted on later. They are made at the architecture stage or not at all.
Sceptre demanded a different way of building weapons
That philosophy shaped Sceptre from the beginning. Sceptre is not simply a longer-range precision-guided artillery round. It is an architecture for precision effects that assumes its own improvement. Its modular design allows propulsion, navigation, seekers, guidance algorithms and payload options to evolve independently without redesigning the entire weapon around each change. That matters most where the threat evolves fastest.
Navigation resilience, electronic protection and guidance performance are increasingly software and sensor problems and software and sensor problems are precisely what an iterative architecture is designed to absorb.
Its development programme reflects exactly the same philosophy. In April 2026, Tiberius Aerospace demonstrated the world's first liquid-fuelled ramjet ignition from a NATO-standard 155mm howitzer, following launch loads approaching 18,000g. That milestone was built on successful fin deployment and stable aerodynamic flight. More recently, Sceptre demonstrated powered acceleration following ramjet ignition, validating another critical stage in its propulsion programme. Each test flight is more than a demonstration. It generates data. That data informs the next configuration. Every iteration makes the weapon better.
But developing Sceptre also exposed something larger. The challenge was no longer simply designing an iterative weapon. It was building an industrial system capable of iterating with it.
Building Sceptre led us to build GRAIL
Traditional defense programmes separate design, production and sustainment into distinct phases managed by different organisations. That model works when weapons are expected to remain largely unchanged for decades. It breaks down when capability is expected to evolve continuously. Developing Sceptre made that impossible to ignore.
Engineering iteration is only half the challenge. The industrial base must be able to absorb change at the same speed. A new guidance algorithm is only valuable if manufacturers can integrate it rapidly. A new seeker matters only if suppliers, production lines and validation processes can move just as quickly. Engineering velocity means little if manufacturing remains static. That realisation led us directly to GRAIL.
GRAIL did not begin as a software platform looking for a defense application. It grew directly out of the challenge of building Sceptre. It connects operational requirements, engineering teams, suppliers and production partners across federated manufacturing networks, allowing validated improvements to move rapidly from design into production without restarting the acquisition process. Rather than treating manufacturing as the final stage of development, GRAIL makes manufacturing part of continuous development. The weapon and the industrial base evolve together.
Why the CRADA matters
That philosophy also explains why our recently announced Cooperative Research and Development Agreement (CRADA) with the U.S. Army Combat Capabilities Development Command Armaments Center represents an important milestone.
A CRADA is neither a procurement contract nor a production order. Instead, it provides a legal framework through which government laboratories and industry collaborate on mutually agreed research, sharing technical expertise, facilities, testing and scientific knowledge to accelerate technology development. In other words, it is designed to accelerate learning.
Its value is not simply access to facilities or expertise. It is the ability to shorten development cycles. Engineers work together. Test data informs the next design. Research feeds directly into subsequent development rather than waiting for an entirely new acquisition programme. That mirrors the philosophy behind both Sceptre and GRAIL. The significance therefore extends well beyond a single artillery programme. It reflects growing recognition that future military advantage will depend less on acquiring finished systems than on creating organisations, industrial processes and partnerships capable of continuously improving them. The fastest learning system, not simply the best initial design, will increasingly determine battlefield advantage.
Measured by iteration velocity
For decades, defense programmes have been judged by delivery milestones. Increasingly, they will be judged by learning velocity. How quickly can software be updated? How rapidly can a new seeker be integrated? How fast can electronic protection evolve? How efficiently can battlefield feedback become an engineering change? How quickly can manufacturing respond? Weapons are ceasing to be finished products.
The industrial age rewarded whoever could build the best weapon. The information age will reward whoever can improve it fastest. That requires more than a different weapon. It requires a different way of designing, testing, manufacturing and partnering. That is why Sceptre was built as an iterative weapon.It is why GRAIL exists. And it is why partnerships such as our CRADA with the U.S. Army Combat Capabilities Development Command Armaments Center matter.
The future of defense will not be defined by platforms. It will be defined by learning velocity.