Here come the Drone Rangers

"The belief in technological superiority is the most dangerous illusion in warfare”

With the Strategic Defence Review (SDR) now expected to be published on Monday, its content and intent are becoming clearer. Unfortunately, despite declarations that “lessons will be learned from Ukraine” it appears that only those lessons which support pre-existing narratives are being embraced. More fundamental insights, such as the enduring necessity of mass, risk being overlooked entirely.

One example, which I wrote about last week, is drone data. It seems even ministers are now advocating a premature shift toward autonomy, driven by a partial understanding of incomplete data.

There is no question that unmanned and autonomous systems will play a vital role in future conflicts and could well be decisive. However, all available open-source data confirms that artillery remains as indispensable as ever. An overemphasis on so-called panacea technologies, at the expense of striking a balanced integration between the new and the proven, ignores Ulysses S. Grant’s timeless reminder that “war never changes.”

These complex systems would be difficult enough to integrate into a well-funded, fully modernised force. Attempting to do so within a military struggling with core fundamentals such as net personnel losses, the looming expiry of PFI contracts and critical equipment nearing obsolescence will be exponentially more difficult than is currently being acknowledged.

With the much-needed increase in defence spending still two years away, there’s a real risk that this fixation on emerging technologies will divert attention and resources from more pressing issues. That’s deeply concerning, especially when contemporary MoD studies contain observations like;

“The scale of the British Army’s and Royal Marines’ “say-do gap” with regards to urban operations is a topic which should be a concern to anyone serious about preparing the UK Armed Forces for modern conflict”

To avoid building on sand, perhaps any uncommitted funds prior to the actual spending uplift should be directed toward addressing the critical but unglamorous shortcomings, rather than chasing the exquisite. Or to put it another way: should we not focus on buying a new iPhone before we start downloading new apps?

Understanding current missteps

“Trust, but verify” Ronald Reagan

Kill chain vs killing: As previously discussed, there is a growing misunderstanding surrounding the role of Uncrewed Aerial Systems (UAS). Whether they are primarily enhancing the kill chain through ISTAR (Intelligence, Surveillance, Target Acquisition, and Reconnaissance), or directly conducting the killing via FPV (First Person View) strikes. The reality, of course, is that UAS are doing both.

However, the misconception that their role is solely kinetic risks skewing priorities — placing UAS above, rather than alongside, artillery. This would improve our ability to "Find" targets, but at the expense of the equally vital "Fix and Strike" functions. If the bulk of the actual killing is indeed still being done by artillery, then the plan to only begin ramping up barrel production in 2027 is simply not good enough.

There’s little value in spotting the enemy if we can’t then target and destroy them effectively. Yet that’s exactly where we’re heading, trading away essential, proven capabilities in exchange for unproven promises. Like selling the cow for magic beans.

Equipment monomania: AI companies will always promote AI. Drone companies will always try to sell drones. 4GD will always advocate for better training. That’s to be expected. But this is exactly why Defence Lines of Development (DLODs) exist – to ensure that products and systems are considered holistically, and to prevent overemphasis on one aspect at the expense of others, the implications of which will be covered in a subsequent article.

The apparent lack of pan-DLOD delivery, in my view, is one of the reasons this current technological monomania has been allowed to take hold.

Another factor is the complexity of these systems, AI in particular. In our experience, understanding of AI within the MoD remains nascent. There also appears to be a sort of corporate embarrassment that prevents the right (often basic, but essential) questions being asked of industry regarding the limitations of AI. If we are to gain a realistic appreciation of what AI can and cannot do, this hesitation must be overcome.

To be clear, autonomy has advanced significantly and will continue to do so. However, our experience suggests that today’s AI would still struggle to reliably distinguish between a British and Russian soldier, especially at the speeds involved in a UAS terminal attack. So, while the idea of a fully autonomous kill chain may have theoretical appeal in linear conflicts, and may tempt a shrinking military, it is far from ready for the complexity of modern, high-density warfare.

Offsetting casualties: The BBC’s 2014 documentary The Lion’s Last Roar insightfully explored the challenges the military faces when it comes to modern civilian tolerance for casualties. In this context, sleek sales pitches promoting fully, or largely, autonomous battlefields can appear appealing, especially when contrasted with the scale of losses seen in Ukraine.

However, these visions are not grounded in current reality. The idea that technology can remove all risk to our service personnel is a dangerous illusion.

Instead of comforting ourselves with high-tech fantasies, we must confront the harder task – educating the public on the true nature of major conflict. Shielding society from that reality with hype and hyperbole does a disservice not just to the Armed Forces, but also to national preparedness.

New operating models: The proposed 20-40-40 doctrine represents a reimagining of force composition, with 80% of capability envisioned to come from autonomous and uncrewed systems. But this concept must be tempered by the reality that such systems could be rendered ineffective by adversary countermeasures such as jamming, deception, cyber attacks, or kinetic counter strikes.

If that happens, the remaining 20%, the conventional force, is likely to still find itself bearing the brunt of combat, including close-quarters fighting, as has so often been the case throughout military history.

No matter how advanced the strategy, it is our shared duty to ensure that this force is trained, equipped, and mentally prepared for the enduring, brutal reality of war at close range.

Issues UAS will have to overcome

“Technology is a useful servant but a dangerous master" Christian Lous Lange

Procurement: It’s easy to lay the blame for the high cost of weapon systems at the door of industry. It’s even easier to compare the price of commercially available drones and imagine replacing a Javelin missile with an FPV drone for just 1% of the cost. But this line of thinking mistakes a symptom for the disease.

Unless the procurement system is reformed before UAS is formally acquired at scale, significant cost inflation is inevitable. Beyond basic materials and manufacturing, suppliers must account for the entire acquisition burden, testing, compliance, integration, and the bureaucratic complexities of becoming an MoD supplier and, in the case of lethal UAS, additional costs such as safety certification, secure storage, and more.

If the MoD treats these systems as generic platforms rather than as munitions with a defined lifecycle, any perceived cost savings will quickly evaporate. It’s not difficult to imagine a scenario where a British-procured lethal FPV drone, once fully onboarded, ends up costing nearly as much as a Javelin missile.

The reality is telling: the average operating profit of four of the world’s largest defence companies is just 9%. This highlights both the inefficiencies in the procurement process and the fact that high prices are often a result of structural issues, not excessive profiteering.

Integration: The long-heralded “digital backbone” remains, for now, a figment of the MoD’s imagination. Without a robust and coherent digital core, it’s difficult to see how the seamless, integrated, and autonomous kill chain, championed by ministers, can become a reality.

If the goal is an integrated system akin to Israel’s “Z-Tube,” then the foundational infrastructure must be built now. Delaying this work risks turning vision into vapour.

A friend recently put it well – it's like giving a complex Lego Technic set to a two-year-old. We need to master the Duplo first, by developing basic digital literacy and interoperability, before attempting advanced integration.

To stretch the introduction’s analogy one step further – we are currently trying to download apps onto a phone that doesn’t exist, over a network that hasn’t been built.

Defence venture complex: I have been a strong advocate for both defence entrepreneurs and emerging technologies for over a decade. However, that advocacy must come with a clear warning – early-stage companies, by nature, tend to be overly optimistic and often underestimate the complexities of scaled manufacturing. Warfare does not tolerate optimism, and soldiers cannot afford to wait for delayed deliveries.

We must build a system that encourages innovation, but it must also rigorously test, challenge, and refine technologies, especially lethal ones, long before they reach operational service. It is neither sustainable nor responsible to spend limited funds on bleeding-edge equipment when existing capabilities still meet end user need.

To that end, I would urge the MoD to establish a well-funded development organisation, akin to the US’s DARPA, that can shepherd promising technologies all the way from TRL 0 to TRL 9 before they are fielded. This would bridge the gap between invention and integration, without pushing unproven systems into service prematurely.

At the same time, private capital must recognise that defence innovation demands a different kind of patience. Returns will take longer than in conventional venture models, and defence leaders must be honest about this timeline. Overpromising commercial outcomes, as has happened recently, only undermines credibility and risks distorting market dynamics.

Creating this ecosystem is of national importance. So too is the growth of our defence industrial base. But placing the burden of panacea solutions on early-stage companies, before their systems are tested, matured, and trialled, will only exacerbate the very problems we are trying to solve.

Supply Chain: If the MoD is truly committed to integrating UAS into future capabilities, it must recognise that the challenges go far beyond the platforms themselves. At present, only 3% to 5% of the cameras required for FPV drones are manufactured in Europe. In the event of a major conflict, supply chain disruption is not a risk, it’s a certainty.

If we intend to rely on these capabilities at scale, then a cross-government strategy to onshore key manufacturing is essential. Equally important is transparency around the origin of components. Without both, we risk building our future defence posture on fragile and foreign-dependent foundations.

Summary

“If I had an hour to solve a problem, I’d spend 55 minutes thinking about the problem and five minutes thinking about solutions” Albert Einstein

Uncrewed Aerial Systems (UAS) and autonomous technologies will undoubtedly play a critical role in any future kill chain, and we at 4GD have been developing synthetic training for both since 2016. However, software and technology alone will not shield British forces from the realities of modern warfare. Nor will they resolve the systemic challenges currently facing the Ministry of Defence. Relying on UAS alone would be like trying to save Theranos with new accounting software.

It’s worth asking whether investing in the fundamentals, such as recruitment, artillery, logistics etc., within the current era of 2.3% to 2.5% of GDP might lay the stable, enduring foundation necessary for effective spending on more complex capabilities as funding approaches (and hopefully exceeds) 3%.

I am, and always will be, a strong advocate for technology and the companies that develop it. But in an era of large-scale, conventional warfare, we must be cautious not to turn AI and autonomy into 2025’s Wunderwaffe or a modern-day Maginot Line.