From Smart Drones to Digital Twins: What Mexico Still Isn’t Using to Fight the Cartels

If the state continues to rely on ground patrols and post-incident responses, the gap will widen.

By Ghaleb Krame

Last May, at the World Police Summit in Dubai, my Australian colleague Dr. Amanda Davies and I presented an academic paper titled Digital Twin Strategies for Countering Cartel Drone Operations: A Policy-Driven Approach. Soon to be published by Springer, the paper outlines a solution that may sound like science fiction—but isn’t. Defense agencies across the United States, Europe, and Asia are already using it: digital twins designed to simulate, predict, and neutralize the threat of drones in hostile environments.

A digital twin isn’t a robot, a drone, or a magical algorithm. It’s a real-time virtual replica of a physical space—say, a border region, a city, or a conflict zone. This technology lets security forces simulate attack scenarios, predict movements, and optimize tactical responses—without having to deploy boots on the ground or wait for a threat to materialize. Siemens uses it to manage smart ports; NASA uses it to model space missions. That same technology can—and should—be deployed to fight cartel drones.

While drug cartels are already using thermal imaging, autonomous navigation, and rudimentary drone swarms, Mexico continues to operate with a fundamentally reactive mindset: random patrols, limited jamming capabilities, and a heavy dependence on traditional intelligence. The technology gap is widening—not due to a lack of resources or talent, but because Mexico still lacks a national strategy to prioritize predictive technologies in its security operations.

This article is not a glimpse into the future. It’s a case for a concrete, tested, and accessible tool that Mexico could be using right now to counter one of the most asymmetrical threats of this century. The cartels are already flying over us. The question is: will we keep staring at the sky, still unsure where the next drone will come from?

2. The Cartel’s Technological Leap: Drones, Autonomy, and Mini-Swarms

Gone are the days when drones were just toys for hobbyists. Since at least 2017, Mexican cartels—especially the Jalisco New Generation Cartel (CJNG)—have integrated drones into their operations for surveillance, smuggling, and attacks. What began as a tool for spotting highway patrols or moving small packages has evolved into weaponized devices fitted with explosives, thermal cameras, and autonomous nighttime flight capabilities (Digital Twin Strategies, 2025).

These drones aren’t just used to drop bombs. They perform ISR (intelligence, surveillance, and reconnaissance), spread propaganda, and assert territorial control. In towns like Tepalcatepec or Chinicuila, drones have been used to intimidate communities, monitor military checkpoints, and film attacks for online dissemination. They often fly below 120 meters—beneath radar detection—and operate without real-time human control, following pre-programmed GPS routes with basic autonomous navigation (Grey Dynamics, 2024).

A key example came in Aguililla, Michoacán, where a coordinated drone attack was reported. It wasn’t a full-fledged military swarm, but it was a functional proto-swarm: multiple drones programmed to converge simultaneously on targets with improvised explosives. This shift—from improvised ambushes to semi-automated, synchronized strikes—signals a strategic evolution (Small Wars Journal, 2021).

There’s also technical refinement. The X account Pernicious Propaganda has documented how CJNG has begun attaching stabilizing fins to their drone-dropped explosives—a feature more typical of ballistic weapons than cartel improvisation. This isn’t random experimentation. It shows learned optimization, possibly shared across cells or supported by outside expertise.

Are cartels using AI? There’s no confirmed evidence of real-time artificial intelligence in cartel drones. But their rapid adoption of thermal modules, extended-range batteries, and pre-programmed navigation suggests they could soon incorporate basic AI functions, such as autonomous flight path correction or target recognition. It's still speculative—but increasingly plausible (Guerra de drones, 2023).

This problem isn’t confined to Mexico. Reports from Arizona Public Media and Grey Dynamics estimate that up to 1,000 drone incursions occur every month along the U.S.–Mexico border. Many involve modified commercial drones used for smuggling, surveillance, or overwhelming defensive systems—flying low, at night, and often in groups to stretch detection capacity (Arizona Public Media, 2024).

This is low-cost, high-adaptability, asymmetric warfare. The cartels are innovating fast. The response can't be more guesswork and checkpoints. It must be predictive intelligence, simulated scenarios, and real-time modeling.

Enter digital twins.

3. SEDENA Has Drones—But Not Predictive Intelligence

Mexico isn’t technologically barren. The country’s Secretariat of National Defense (SEDENA) operates several advanced drones, including the Hermes 450 and 900 (built by Elbit Systems) and the S-45 Baalam, a Mexican-made model from Hydra Technologies. These UAVs have solid flight endurance and high-resolution sensors. The Hermes 450 can fly for 17 hours and carry up to 150 kg; the Hermes 900 goes up to 30 hours and 300 kg and was even deployed for search-and-rescue operations during the 2024 Brazil floods. The Baalam has been used for maritime surveillance and disaster response with up to 12 hours of autonomy and synthetic-aperture radar (Digital Twin Strategies, 2025).

But technical capacity isn’t the same as tactical foresight.

Despite these platforms' potential, there’s no public evidence that SEDENA has integrated them into predictive modeling systems or digital twin environments. They’re largely used for pipeline surveillance and ad hoc ISR operations—not as nodes in a broader, simulation-based architecture. Unlike U.S. Customs and Border Protection (CBP), which is already testing pre-trained AI models like YOLOv8 for real-time drone detection, Mexico shows no indication of running Monte Carlo simulations to predict cartel drone behavior (Jocher et al., 2023; CBP, 2022).

Unlike countries such as Israel, South Korea, or the United States—where digital twin simulations are already being used to test and refine military responses—Mexico remains stuck in operational, not strategic use of its drone fleet. There is, at least publicly, no integrated system that merges geospatial data, predictive AI, and field operations.

Using drones without simulation is like patrolling without a map. The hardware exists. What’s missing is the virtual intelligence infrastructure to turn each flight into a source of strategic foresight.

4. What Is a Digital Twin—And Why It Matters for Security

A digital twin is not a futuristic abstraction. It’s a real-time, data-fed simulation of a physical environment. Already used across aerospace, logistics, and critical infrastructure, it has powerful applications in national security.

NASA uses digital twins to model flight conditions and prevent failures before they happen. Siemens applies them to optimize port operations. The U.S. Army Futures Command uses synthetic environments to simulate thousands of drone attack scenarios—testing different defense strategies without risking lives (U.S. Army Futures Command, 2024; Siemens, 2023).

In Europe, the Copernicus platform uses digital twins to model irregular border crossings and optimize sensor placements. The results: faster response times, better coverage, fewer blind spots (Copernicus, 2025).

For Mexico, digital twins could simulate cartel drone routes, radar blind zones, interdiction windows, and the effectiveness of countermeasures—all based on real terrain and OSINT data. Combined with object detection models like YOLOv8 and satellite inputs, the system could identify potential threats before they materialize, helping law enforcement allocate resources with greater precision (Digital Twin Strategies, 2025).

This isn’t about reacting faster. It’s about predicting smarter.

And above all, this isn’t science fiction. The tools are already here: open software, public datasets, proven platforms. What’s missing isn’t the tech. It’s the political will.

5. Why Hasn’t Mexico Made the Leap?

Mexico has drones. It has trained operators. It has access to geospatial data and affordable detection technologies. What it lacks is a national framework for predictive security.

Agencies like CBP are already integrating acoustic sensors, tower surveillance, and AI detection tools into predictive platforms. Mexico is still relying on ground patrols and line-of-sight interdiction. The gap is stark: the U.S. simulates threats before they occur. Mexico reacts when it’s already too late (CBP, 2022).

Countries like South Korea and the UK are moving forward with synthetic training environments for asymmetric threats. The Copernicus system in the EU is not a prototype—it’s operational. In contrast, Mexico lacks any agency capable of consolidating digital twins, AI, OSINT, and drone operations under a unified doctrine (Copernicus, 2025; Siemens, 2023).

Worse still is institutional fragmentation. SEDENA, the National Guard, and the intelligence community operate in silos—with disjointed protocols, separate chains of command, and priorities too often shaped by electoral cycles rather than strategic necessity. Even promising homegrown projects like the S-45 Baalam remain disconnected from simulation or predictive systems.

The result isn’t just inefficiency. It’s vulnerability.

The cartels aren’t waiting for the government to coordinate. They’re already there—and airborne.

6. Five Concrete Steps to Anticipate the Drone Threat

Mexico doesn’t need to start from scratch. The technology exists. The models are proven. The data is public. Here’s what the country can do—now:

• Deploy digital twins in critical regions.

Areas like Tierra Caliente, the Isthmus of Tehuantepec, and border corridors near the Rio Grande can be virtually replicated using NASA, USGS, or Copernicus data to simulate cartel activity and optimize response.

• Integrate open-source AI for early detection.

Models like YOLOv8—already field-tested by CBP—can detect drones with 85% accuracy and millisecond response time. With basic training, these tools can be localized.

• Feed the system with OSINT.

Publicly available videos, citizen reports, satellite patterns, and cartel propaganda can help adapt simulations in real time and identify tactical shifts.

• Legally simulate countermeasures.

Technologies like GPS spoofing or RF jamming face legal hurdles. But simulating their effects allows Mexico to test these tools safely, assess risks, and develop legally viable deployment protocols.

• Create a binational tactical intelligence hub.

Drones don’t respect borders. A U.S.–Mexico predictive center for OSINT and simulation would help standardize alerts, align training, and coordinate interdiction—without compromising sovereignty.

7. Conclusion: What Isn’t Simulated Can’t Be Controlled

In aviation, medicine, and now in security, the difference between failure and prevention lies in simulation. No one expects a pilot to fly without a simulator. So why do we expect our security forces to face aerial threats without one?

Mexico doesn’t need a tragedy to act. The tools exist: public datasets, open-source models, and proven digital twin platforms. What’s missing is not hardware—but strategy.

The cartels already operate low-cost drones with autonomous navigation, enhanced explosives, and rudimentary swarming capabilities. If the state continues to rely on ground patrols and post-incident responses, the gap will widen.

This isn’t science fiction. This is a solvable institutional lag—if the political will is there.

The cartel war has taken to the skies. Mexico’s security future won’t be decided only on the ground. It will be shaped in the virtual battlespace.

Let’s just hope simulation remains the domain of artificial intelligence—

and not of political theater.