Identifying Capability Gaps in Defense Technology

Most military organisations are investing heavily in technology. New sensors, drones, AI capabilities, communications systems, data platforms. The difficult question is whether we are preparing those capabilities for the conditions they are most likely to face. Much of modern defence still assumes reliable communications, available positioning, trusted data, predictable logistics, and stable information flows. Yet recent conflicts continue to show an environment where communications degrade, GPS becomes contested, data is manipulated, adaptation cycles are measured in weeks, and systems are continuously exposed to electronic warfare, cyber pressure, and information operations. The challenge is not that technology is failing. The challenge is that many of the assumptions supporting technology are increasingly becoming targets themselves. In the first full article of the Commander’s Handbook – Cutting Through the Noise series, I explore the gap between the war many systems are designed for and the war they are likely to encounter. The article looks at: • why modern conflicts are exposing hidden dependencies • what Ukraine reveals about adaptation, drones, and electronic warfare • why drone warfare is far more complex than the public narrative suggests • how incidents now spread across military, cyber, information, infrastructure, and political domains within hours • what commanders should be asking before the next crisis rather than during it The purpose is not to predict the next war. It is to challenge the assumptions we may be carrying into it. Article below. #Defence #MilitaryInnovation #CommandAndControl #OperationalReality

Most AI signal processing pilots fail. 𝗡𝗼𝘁 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝘁𝗵𝗲 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗶𝘀𝗻'𝘁 𝗿𝗲𝗮𝗱𝘆. 𝗕𝗲𝗰𝗮𝘂𝘀𝗲 𝘁𝗵𝗲 𝗼𝗿𝗴𝗮𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝗶𝘀𝗻'𝘁. After analyzing 𝟮𝟬𝟬+ AI initiatives across telecom, defense, and industrial IoT, the pattern is unmistakable: → Organizations scoring "operational" on 𝟱+ readiness dimensions: 𝟴𝟵% production success rate → Organizations with 𝟯+ "aspirational" dimensions: 𝟭𝟮% production success rate 𝗧𝗵𝗲 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝘄𝗮𝘀 𝗻𝗲𝗮𝗿𝗹𝘆 𝗶𝗱𝗲𝗻𝘁𝗶𝗰𝗮𝗹 𝗮𝗰𝗿𝗼𝘀𝘀 𝗯𝗼𝘁𝗵 𝗴𝗿𝗼𝘂𝗽𝘀. The difference was 𝗿𝗲𝗮𝗱𝗶𝗻𝗲𝘀𝘀 — assessed 𝗯𝗲𝗳𝗼𝗿𝗲 vendor selection. We built a framework around the six dimensions that predict production outcomes: 𝟭. 𝗗𝗮𝘁𝗮 𝗜𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 𝗠𝗮𝘁𝘂𝗿𝗶𝘁𝘆 Streaming pipelines vs. batch extraction. Most organizations overestimate where they are. 𝟮. 𝗦𝗲𝗻𝘀𝗼𝗿 𝗜𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻 𝗗𝗲𝗽𝘁𝗵 ⚡ API-accessible data vs. manual export workflows. 𝗧𝗵𝗶𝘀 𝗼𝗻𝗲 𝗸𝗶𝗹𝗹𝘀 𝗺𝗼𝗿𝗲 𝗱𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁𝘀 𝘁𝗵𝗮𝗻 𝗮𝗻𝘆 𝗼𝘁𝗵𝗲𝗿. 𝟯. 𝗥𝗲𝗮𝗹-𝗧𝗶𝗺𝗲 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴 𝗖𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝘆 Edge vs. cloud architecture. The right answer depends on latency requirements and connectivity constraints. 𝟰. 𝗢𝗽𝗲𝗿𝗮𝘁𝗼𝗿 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄 𝗜𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻 Is AI embedded in how operators work — or running in a separate tab no one opens? 𝟱. 𝗥𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘆 𝗖𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲 𝗣𝗼𝘀𝘁𝘂𝗿𝗲 FedRAMP, ITAR, CMMC. In defense and critical infrastructure, 𝘁𝗵𝗶𝘀 𝗶𝘀 𝗮 𝗴𝗮𝘁𝗲, 𝗻𝗼𝘁 𝗮 𝗰𝗵𝗲𝗰𝗸𝗯𝗼𝘅. 𝟲. 𝗖𝗿𝗼𝘀𝘀-𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝗹𝗶𝗴𝗻𝗺𝗲𝗻𝘁 IT/OT convergence maturity. Where most organizations have the 𝘄𝗶𝗱𝗲𝘀𝘁 𝗴𝗮𝗽 between where they think they are and where they actually are. Organizations that ran this assessment 𝗯𝗲𝗳𝗼𝗿𝗲 selecting a vendor: → Redirected failing initiatives → Avoided 𝗿𝗼𝘂𝗴𝗵𝗹𝘆 $𝟮𝗠–$𝟴𝗠 in sunk costs → Cut time-to-production by an average of 𝟭𝟰 𝗺𝗼𝗻𝘁𝗵𝘀 One defense contractor was 𝟲 𝘄𝗲𝗲𝗸𝘀 𝗳𝗿𝗼𝗺 𝗮 $𝟰𝗠 𝗱𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁 when the assessment flagged a fatal compliance gap. Redirected. Saved. Now at L4 Operational 14 months later. Comment 𝗔𝗦𝗦𝗘𝗦𝗦 for the full framework, scoring rubric, and case studies.

𝗡𝗔𝗧𝗢 𝗶𝘀 𝗻𝗼𝘁 𝗮𝘀𝗸𝗶𝗻𝗴 𝗳𝗼𝗿 𝗮𝗻𝗼𝘁𝗵𝗲𝗿 𝗹𝗼𝗻𝗴-𝗿𝗮𝗻𝗴𝗲 𝗱𝗿𝗼𝗻𝗲 NATO Allied Command Transformation has launched a public Innovation Challenge seeking systems capable of persistently denying enemy airfield operations—a requirement based explicitly on Ukrainian battlefield experience. As William Lawson notes in The National Interest, Russian tactical aviation continues operating from secure rear-area airfields beyond the reach of most conventional Ukrainian strike assets, launching guided bombs, cruise missiles and other stand-off weapons against Ukrainian forces, infrastructure and cities. Ukraine can already strike individual airfields, aircraft and support facilities. The capability gap identified by NATO is different: Ukraine cannot yet generate sufficient mass, persistence and electronic-warfare resilience to suppress several defended airfields—or multiple critical points within one airfield—continuously. The challenge therefore seeks more than another one-way attack drone. Proposed systems must operate at ranges exceeding 150 kilometres, function in GPS-denied and heavily jammed environments, work in all weather and attack aircraft, runways, fuel storage, ammunition facilities and ground-support infrastructure. NATO is open to uncrewed aircraft, autonomous or semi-autonomous munitions, swarms, alternative overland delivery systems and hybrid architectures. Continuous operator control is unacceptable without autonomous fallback, while AI-assisted target acquisition, rapid software modification and simultaneous engagement of multiple aim points are specifically emphasized. The timetable is equally revealing. Systems requiring more than 12 months to reach the field are excluded, technology readiness must already exceed TRL 5, and delivery of a minimum viable capability within six weeks is preferred. Up to ten finalists will be invited to a pitch event provisionally planned for Warsaw in September. This is an Innovation Challenge rather than a procurement commitment, but the requirement itself matters. NATO is publicly acknowledging that intercepting Russian bombs and missiles after launch must be complemented by suppressing the airfields that generate them. Persistent airfield denial does not necessarily require permanently destroying an entire base. It means repeatedly closing runways, damaging fuel and weapons infrastructure, threatening parked aircraft and striking repairs quickly enough that Russian sortie generation becomes unsustainable. That is the transition from occasional deep strikes to an operational campaign: attacking not only what flies from a Russian airfield, but Russia’s ability to keep that airfield functioning. 𝘛𝘩𝘦 𝘨𝘰𝘢𝘭 𝘪𝘴 𝘯𝘰𝘵 𝘵𝘰 𝘩𝘪𝘵 𝘢𝘯 𝘢𝘪𝘳𝘧𝘪𝘦𝘭𝘥 𝘰𝘯𝘤𝘦; 𝘪𝘵 𝘪𝘴 𝘵𝘰 𝘮𝘢𝘬𝘦 𝘦𝘷𝘦𝘳𝘺 𝘳𝘦𝘱𝘢𝘪𝘳 𝘵𝘦𝘮𝘱𝘰𝘳𝘢𝘳𝘺 𝘢𝘯𝘥 𝘦𝘷𝘦𝘳𝘺 𝘯𝘦𝘸 𝘴𝘰𝘳𝘵𝘪𝘦 𝘶𝘯𝘤𝘦𝘳𝘵𝘢𝘪𝘯.

"Nations that control advanced AI systems will shape economic competitiveness, military capability, and the terms on which governments function for decades. The asymmetries created by unequal access to frontier AI—and the dependencies created by concentrated control over its inputs—have elevated AI to the center of great-power competition. Governments worldwide are now pursuing “AI sovereignty”: the ability to develop and deploy frontier AI on their own terms and sustain those capabilities under external pressure. This report provides a comprehensive asset-level, cross-country analysis of AI sovereignty. We then evaluate each country’s autonomy (control over AI capabilities) and resilience (ability to sustain capabilities under shocks) and derive strategic recommendations for U.S. policymakers. Critical Findings Both the U.S. and China offer partner countries “vendor-chaperoned sovereignty”—the promise of AI capability under conditions that embed dependence. The nature, risks, and strategic implications of this export competition are analyzed in the Introduction. Four chokepoints define the global AI supply chain. ASML[1] ’s 100% monopoly in extreme ultraviolet (EUV) lithography, TSMC’s dominance in advanced logic manufacturing, the Korean duopoly in high-bandwidth memory, and China’s control over critical minerals processing (60–70% of rare earth mining and 80–90%+ of rare earth processing, ~98% gallium, ~60% germanium) create interdependencies that no single nation can escape. High-bandwidth memory is the tightest near-term bottleneck. All high-bandwidth memory (HBM) production for 2025–2026 is sold out[2] . Micron[3] can meet only ~55–67% of its 2026 demand[3] . Korean producers control over three-quarters of global supply. Every advanced AI accelerator—from Nvidia, AMD, or emerging designers—requires HBM that only Korean and U.S. firms produce. The frontier model capability gap is narrow. Chinese open-source models have reached global 5 of 87 Visions of Sovereign AI penetration of ~30%. The AI chip performance gap remains significant[4] but is narrowing. Huawei’s Ascend 910C achieves 60% of Nvidia H100 inference performance[5] . The semiconductor manufacturing equipment gap—China’s most critical constraint—is measured in years, not months, and represents the most durable U.S. leverage point. SMIC has produced 7nm chips using deep ultraviolet (DUV) workarounds, but investment through Big Fund III ($47.5 billion) is accelerating domestic alternatives. Critical minerals represent the U.S.’s most acute vulnerability. The U.S. imports 100% of its gallium[6] . China’s ability to simultaneously restrict gallium, germanium, rare earths, and graphite could severely disrupt semiconductor manufacturing and data center scaling. China’s temporary export control suspension expires November 2026 "

𝗔𝗿𝗲 𝗪𝗿𝗶𝘁𝗶𝗻𝗴 𝗥𝗲𝗾𝘂𝗶𝗿𝗲𝗺𝗲𝗻𝘁𝘀 𝘁𝗵𝗲 𝗕𝗶𝗴𝗴𝗲𝘀𝘁 𝗙𝗮𝗶𝗹𝘂𝗿𝗲 𝗣𝗼𝗶𝗻𝘁 𝗶𝗻 𝗠𝗶𝗹𝗶𝘁𝗮𝗿𝘆 𝗔𝗰𝗾𝘂𝗶𝘀𝗶𝘁𝗶𝗼𝗻? In many defence organisations, poorly written requirements are one of the biggest failure points in the entire acquisition process. Cost overruns, delays and capability gaps often trace back to unclear, incomplete or unrealistic requirements set at the start. In my 20+ years in military procurement, I’ve seen most requirement documents built around technical specifications of a desired solution rather than the operational effects and capabilities actually needed. This reversed logic is a key reason why acquisition so often struggles. Another recurring issue is the lack of prioritisation. Without a clear hierarchy between capabilities must-have, nice-to-have and luxury features, organisations tend to chase a “jack-of-all-trades” device – expected to do everything, but optimised for nothing. The result is predictable: • cost overruns • schedule slips • and a system that fails to satisfy the end user A strong capability-based approach, combined with disciplined prioritisation, would prevent many of these pitfalls. Take infrared observation devices as an example. A fundamental capability question is: How quickly do you need to operate the device? This drives key requirements: • Startup time – seconds or minutes? • Accessibility and size – on the vest for immediate use, or in the backpack? If you need a small, rapidly accessible device with very fast startup, something like the Pixel on Target VooDoo-R fits in a mag pouch and starts in under five seconds – but you “only” get an IR channel plus range finder/target locator. If you need multiple channels, you must again prioritise capabilities vs. size and weight. For example: • Safran Moskito TI: 3 channels (direct optics, LWIR, low light) and long-range target location – but larger and heavier. • Elynxo Virtuose: smaller, faster startup, 2 channels (direct optics, LWIR) and shorter range target location – a compromise between versatility and portability. The lesson: if requirements are written as a shopping list of technical features, without operational context and prioritisation, procurement will almost always pursue a “do it all” system that ends up too big, too costly, too slow or too complex for real-world use. With more than 100 successfully executed projects at Special Forces Command, I can support you in defining and refining clear, capability-based equipment requirements. Don’t hesitate to contact me. #Dashuyn #DefenseConsulting #DefenseProcurement #CapabilityDevelopment #DefenseInnovation #MilitaryAcquisition #MilitaryRequirements #SwissArmedForces #SpecialForces #PixelOnTarget #VoodooR #Safran #MoskitoTI #Elynxo #Virtuose

The U.S. defense industrial base is struggling to keep up with global demand, exposing alarming weaknesses in its ability to produce weapons and ammunition at scale. This shortfall has been highlighted by the war in Ukraine, where the U.S. has failed to produce weapons fast enough to supply its allies without depleting its own stockpiles. A Lagging Production System The production challenges stem from decades of underinvestment and reliance on outdated infrastructure. Many U.S. weapons production facilities, like the Iowa plant producing 155-millimeter artillery shells, date back to World War II and lack the capacity to scale output rapidly. Prior to Ukraine’s invasion, the U.S. manufactured roughly 14,000 shells per month. Today’s demand exceeds 80,000 per month, far surpassing the system’s capabilities. Key Problems Include: 1. Aging Infrastructure: Factories are ill-suited to meet the demands of modern warfare, relying on outdated processes and equipment. 2. Workforce and Supply Gaps: Skilled labor shortages and fragile supply chains have further delayed production. 3. Private Sector Dependence: The Pentagon’s reliance on private contractors has introduced inefficiencies in scaling operations during emergencies. Implications for U.S. Military Preparedness The inability to sustain high levels of production raises concerns about the U.S. military’s readiness for a prolonged, high-intensity conflict. Experts warn that a war with a peer adversary, such as China, would quickly exhaust existing stockpiles and overwhelm the production system. Unlike small-scale conflicts, great-power competition requires both advanced technology and the capacity to mass-produce ammunition and equipment. Addressing the Crisis To close this critical gap, U.S. defense officials are pursuing reforms to modernize the defense industrial base: • Infrastructure Investments: Expanding and updating munitions plants to increase production capacity. • Supply Chain Diversification: Reducing reliance on single suppliers for critical components. • Allied Cooperation: Collaborating with NATO and partners to share production and supply responsibilities. A Wake-Up Call The war in Ukraine has underscored a harsh reality: America’s military dominance depends not only on cutting-edge technology but also on industrial resilience. Without immediate reforms, the U.S. risks being unprepared to arm itself for the conflicts of the future. The failure to address these vulnerabilities could have devastating consequences in an era of rising global threats.

There is a story in the news this month about a Chinese aerospace engineer who, over four years, persuaded researchers at NASA, the US Air Force, the Navy, the FAA, and faculty at a number of American universities to send him export-controlled software. He worked at one of the state-owned conglomerates whose name appears regularly in Western sanctions documentation. He used impersonation. The campaign ran on consumer email accounts. It was uncovered by a tip, not by a system. The instinct in the cybersecurity press is to read this as a phishing story. It is not a phishing story. It is a visibility story. The targeted institutions had no structural way to connect a request from a plausible-looking academic email to the institutional behavior of an adversary's defense-industrial system across hundreds of subsidiaries, personnel flows, procurement priorities, and front structures. The pattern of operations of which this engineer's campaign was one instance has been visible in the aggregate for years. No platform serving these institutions was looking at the aggregate. The pieces sat in open record. No one held a frame in which they composed. This is the gap I think about. The cost of incomplete visibility is not paid in headlines. It is paid in intellectual property that walks out the door, four years at a time, until a tip arrives. It is paid in supply chain decisions made on screening tools designed to confirm what is named, not to surface what is structural. It is paid in procurement frameworks that assume a counterparty's declared identity is its operational identity. The engineer is not an anomaly. He is a representative of a category. Generalist platforms cannot see the category because their architecture optimizes for breadth across all adversaries. Aggregator products cannot see the category because their work ends at the registry. The category is visible only to a layer purpose-built for the ecosystem in which it operates. What I have founded RedRadar for, is to provide the West with the analytical layer this category requires. Not a tool. A capability. Built specifically for the ecosystem it is meant to see, for the institutions whose job it is to see clearly and defend accordingly.

CRITICAL MINERALS & DEFENDING THE U.S. - These Materials Could Cripple America’s Defense Industrial Base - A Practical Test for Material Chokepoints - Not every “critical” material is an urgent problem. A material chokepoint exists when five conditions coincide: high concentration in mining or refining, direct defense criticality, low substitutability without performance loss, long, capital-intensive build times, and recent signs of policy leverage such as export licenses, price manipulation, quotas, or prohibitions. - How vital is each material to U.S. defense, and how much leverage does China wield? - GRAPH: Figure 1. This chokepoint matrix flags where a shortfall would ripple through U.S. defense programs the fastest and where the weaponization of supply chains by an adversary would cause major disruptions. It synthesizes public sources (i.e., U.S. Geological Society, International Energy Agency, and export-control reports). Axes are ordinal 0-12 and convey relative differences (not metric). - First-tier risks are gallium, the battery-chemicals chain, tungsten, and graphite. A second tier includes titanium sponge, germanium, antimony, indium, magnesium, and molybdenum, with nitrocellulose and finish-line capacity in specialty steels and aerospace aluminum requiring near-term hedges because surge capacity is slow to add. - NOTE: Crucially, most chokepoints rarely happen at the actual mine. Most problems occur in mid- and downstream steps such as refining, separation, smelting, high-purity processing, alloying, component manufacture, and device-grade finishing. Ore alone does not deliver security. Rather, mid- and downstream control does. - Two pathways close gaps. 1. First, build at home - where chemistry, processing and finishing are the binding constraints rather than geology. This includes materials such as lithium-ion battery chemistry, coated and synthetic graphite anodes, nitrocellulose for propellants, semiconductor wafer and epitaxy capacity, infrared-optics finishing cells, tungsten-carbide recycling, and the finishing lines for specialty steels and aerospace aluminum. - Expanding these capabilities at home reduces exposure to Chinese leverage, grows the U.S. industrial base, and ensures that the finishing steps most critical to defense remain under American control. 2. Second, hedge upstream exposures with allies where ore, smelting, or primary refining are concentrated abroad. - Examples include structured offtake and investment, "U.S. recycling," and interim tolling until U.S. lines qualify. - For each material, align remedy to the chokepoint: If it is chemistry or device-grade finishing, build U.S. lines with price floors and multiyear offtake. If it is about upstream and feed gaps, use allied offtake and tolling as U.S. finishing ramps. When policy leverage appears, apply the playbook. - https://lnkd.in/dr8hgACm

Here's a story that will change how you think about security visibility: In WW II, airplane mechanics were required to mark every bullet hole when damaged planes returned to base. They created detailed diagrams showing where planes got hit most often (wings, tail sections, fuselage). But here's the twist: The areas with NO bullet holes were actually the most critical. Why? Because planes hit in those spots never made it home. This is exactly what's happening in your security program right now. Your SIEM shows thousands of alerts. Your vulnerability scanner flags hundreds of issues. Your cloud security platform reports misconfigurations daily. But what about the systems that AREN'T reporting? The endpoints that went silent? The cloud resources that slipped through the cracks? The legacy systems that were never properly instrumented? The most dangerous security question isn't "What threats did we detect?", it is – "what threats are we completely blind to?" I recently spoke with a CISO managing $180B in critical infrastructure. His approach? He doesn't want vendors to show him what they're detecting. He wants them to map his blind spots. He wants to see the security equivalent of those empty spaces on the WW II damage reports. Three questions every security leader should ask their vendors, and this is what we at Leen are helping answer: 1) What systems in my environment are you NOT monitoring? 2) Where do your detection capabilities have natural limitations? 3) How do you help me identify coverage gaps, not just security events? The scariest part of your security program might be the silence, not the noise.

Powerful analysis with policy recommendations. Blocking adversary access to tech is only one initial step, as is on-shoring ME production. "Knowledge power has two essential elements: the ability to innovate and the ability to anticipate. The first relates to a country’s capacity to produce and harness technological breakthroughs. The second has to do with intelligence. Part of this work fits into the traditional mission of U.S. spy agencies, which are tasked with discovering the intentions and capabilities of foreign adversaries to threaten U.S. interests. As the boundaries between domestic industry and foreign policy blur, however, intelligence agencies also need to help the government understand the implications of technologies developed at home." "U.S. policymakers need a new playbook that will help them assess, enhance, and use the country’s knowledge power.The first step is developing intelligence capabilities to gauge where the United States is ahead in emerging technologies and where it is behind, and to determine which gaps matter and which do not. The Pentagon has legions of analysts comparing U.S. and foreign military capabilities, but no office in the U.S. government does the same for emerging technologies. This needs to change. The Office of the Director of National Intelligence has already begun building stronger relationships with companies and universities to gain insight into U.S. technological developments. These efforts must be institutionalized, with channels to share expertise faster and more frequently. To spur progress, Congress should hold annual technology net assessment hearings with intelligence officials and academic and industry leaders." Casey Blackburn