Red Team Shield, S.L.

In the dual defense domain, we need to learn many lessons from closely related fields such as infrastructure protection. For instance, 3D modeling and AI-assisted anomaly detection are fully mature for generation and power grid maintenance, saving billions per year in technical losses and power outages. During my presentation of Red Team Shield, S.L. at #NATOTIE24, I supported the idea of data as the main Western C-UAS asset. Data is the asset that is most unaffordable and unattainable for non-state actors and many non-allied actors. The future Sapient-based STANAG 4869 is going to be a critical milestone, but more effort is needed in this direction. Precise modeling of a protected area is going to increase the effectiveness of the operational picture. If the protected asset is permanent, a time-based model will support anomaly detection. And anomaly is the name of the game: a hostile UAS is a set of sensor data that does not match the historic dataset. While C-UAS defense demands a very specific set of capabilities, the urgency is so high that reinventing the wheel is an actual and serious threat to C-UAS systems updates. The C-UAS space needs to leverage as much prior art and mature assets as possible in order to catch up with the disruptive evolution of hostile drones.

I have a worrisome hypothesis about public view of drone attacks: people have get used to it and it happens elsewhere. During last NATO TIE24, some conversations happened everywhere, every time. "We all are waiting for something bad to happen". But there are drone-related bad things happening everyday. Thousands, actually. And thousands of daily events, some of them converted into social media, have transformed drone attacks into an unremarkable event from the outside. Novelty effect is completely wasted. And since such daily events are happening outside, the idea is cimented more and more: daily bad things happen where bad things happen. Decision makers are not common people by definition. However, they live and work inside a cultural environment, both local (professional) and general. And it is really difficult to remain isolated of this far+daily nature of drone threat and thus prioritize the potential threat over what has happened in the developed countries. Finally, when factual reality overrides this cultural meaning and far and daily is substituted by here and now, it's not so easy to take measures with time margin. I hope that some decision makers will remain isolated from this cultural trend, prioritize and buy us all some time.

Here we are. I worked for several years on projects related to energy distribution. Since 2018, I have been talking with several stakeholders about the ease of attacking energy distribution substations with drones. Actually, explosives are not the only option. It can happen anywhere. And the impact on communities can be dramatic. Critical infrastructure in even slightly contested regions will have to adapt to this threat, pronto. https://lnkd.in/d-G7UARk

One of the constraints on accelerating the evolution of FPV drones used in the Ukrainian conflict stems from one of their greatest virtues: the use of standard civilian components. Whether it's the use of new frequency bands, video signal scrambling, or especially the employment of Edge AI resources for navigation in GNSS-denied environments or autonomous terminal guidance, most of the main advances of 2023 and 2024 are based on civilian components. The problem lies in that if the adversary manages to obtain a sufficiently intact device, reverse engineering work is facilitated by the use of standard components which, not being general-purpose, are not easy to obfuscate in code. When the adversary completes a reverse engineering task, the result is that they have saved time in developing a similar system or, worse, gained access to a resource they hadn't imagined or that was out of reach due to lacking a critical element. In other words, the frenzied pace of evolution in microdrone components and systems isn't as fast as it could be. On the other hand, the incentive to evolve these systems can be described as a (Alice in Wonderland) Red Queen's race: if one doesn't evolve quickly, falling behind has catastrophic consequences for either side. There's another more worrying scenario: a technologically highly capable adversary. China would be the perfect example, but not the only one. We must not forget that the evolution of microdrones doesn't require enormous and exquisite technological or human resources. Once the possibilities have been demonstrated by example, the incentive to develop these capabilities in a sustained manner is evident. If said adversary spends quarter after quarter on iterative development and concerted responses to different needs, in a relatively short time they can acquire an arsenal: * With completely unprecedented capabilities and without an adequate response from their adversaries. * Oriented towards mass production yet modular and adaptable. * Deployable both massively and sustainably over time. The response to this scenario is frankly difficult. It's not obvious how defensive means can acquire this pace of evolution and adaptability. In all likelihood, data is at the center of this response, but it will possibly require priority and investments far superior to those being employed by who is developing vectors and effectors based on microdrones. All in all, not responding is not an option. Time will pass for everyone, and each party will reap what they sow and at the pace they sow it.

Finally a new generation of Ukrainian FPV drones is maturing. 10 inches plus, mature and capable dropper mechanism, pivoting camera for dropping when hovering (or in certain case for dive bombing), and a protected GPS antenna. Maths are going to be something. Cost is x 2, but if lifespan is more than 8 missions on average, things are going to change. Forget about that specific dropper. Other droppers with unitary charges are already having serious impact. And remember: high gain antennae has nice toroidal shapes. If new queen hornets and their cousins drop payloads from certain altitudes, EW is going to be seriously degrades. And of course, such drones are going to be copied and used outside of Ukraine Lucas Martín Serrano Samuel M. Diego Bueno Pérez https://lnkd.in/ds7YnUaX

Data is the most important foundation of counter-drone development. Now that awareness of the drone defense gap is widespread, NATO partners need to align their vision and roadmap. Vendors have to adapt quickly, change their pace, and realign priorities, putting affordability and scalability at the forefront. Many stakeholders are not yet fully aware of what the SAPIENT protocol actually means for our future. The threat is so dire that competition needs to be strongly balanced with operational cooperation. There is simply too much at risk.

This video from LlamaIndex is raw gold or whatever other suitable metaphor. While actually there is no almost naive RAG in production, nevertheless the masterclass summarizes in 16 min what can be done in order to improve results parting from it. I miss the combination between agents and knowledge graphs, specially because KG should not be treated as a function call, tool, etc., but a happy orthogonal solution to RAG for the knowledge building problem. Yes, these are the days. https://lnkd.in/dBy9aG9v

These are incredibly powerful, flexible, affordable and NEUTRAL resources. In the wars of the hackers, these resources are going to be integrated into both vectors and effectors. Just predict a tank, APC, logistics truck, etc with a microcontroller trained using an LLM is beyond craziness. Entrance barrier is ridiculously low right now and it's going to tend to zero during this year.

La eclosión de los drones comerciales letalizados como elemento de disrupción para la seguridad y defensa demanda nuestra atención en occidente tanto desde Fuerzas Armadas, como desde Fuerzas y Cuerpos de Seguridad del Estado, como de la Industria. Responder a tiempo no es opcional

Drone Swarms as Cultural Artifacts Two days ago, I made some comments about the priority and centrality of drone swarms as a threat https://lnkd.in/dT-BwxgP No matter what actual events show every day in more and more parts of the world, unseen swarms are threatening everything all the time. Besides the development of expensive systems capable of dealing with such a threat, drone swarms have been dominating the domain debate for more than ten years in a row. Year after year, we have observed worrisome developments in drone technology, BUT THE SWARM. And here we are: striving to protect against swarms while protection against remotely piloted UAS is not good enough. Attackers have too much of an advantage; they have changed the course of war in Ukraine, and the Olympics are this summer. I don't want to think in conspiracies. They are quite resilient to Occam's razor. Instead, besides raw popularity and collective think inertia, I would like to explore a different path: cultural artifacts. You see, dogfight air combat was real for just less than 20 years. It was THE air combat during World War I and several minor conflicts right after it, but during the middle 1930s, it was being abandoned in favor of energy combat, boom and zoom, and other non-visually appealing combat modes. Not many years after the Second World War, AA missiles were created and quickly developed, and we can confidently say that during the last 50 years, missiles have been the kings of air combat, following known stats and figures. The problem with air combat is that it was visually narrated during and right after its creation using the then also young cinema. Western and world public was captivated by mesmerizing and exciting air combats between chevaliers of the air that turn on a dime looking for the tail of their adversaries. No matter how terrible actual WWI air combat was, the film portrait of air combat was engraved in stone, so to speak. Decade after decade, although actual air combat evolved, it was portrayed in films as dogfight. Right up until 2023. Now, it's not so easy to trace how drone swarms acquired popularity. Slaughterbots activists campaigned, sure, and it looked like an asset only accessible to main powers. It was more or less the direct translation of shock-and-awe capabilities of the nineties, and also a very striking picture. If you want to portray the threat posed by small drones on a computer display, it's better to make them fly ominously in flocks against their objectives. Like killer bees or the birds of Hitchcock. Nowadays, it's not debatable that swarms have captured the imagination. While not operative, we talk about them all the time and prepare against them. However, there is a decisive difference: dogfights were a cultural artifact AFTER they happened, while swarm attacks have not happened yet.