It was tested during the fifth iteration of Flytrap, which was initiated by the US and UK in 2025. This iteration explored force-on-force experimentation as well as a range of counter-drone technologies and 3D printing.<\/p>\n
Calibre understands that approximately 200 total drone flights were conducted during the exercise, and that US troops launched Helsing\u2019s HX-2 drone 17 times. The drone recorded 15 successful target strikes (direct hits) and two near-misses, which translates to a success rate of over 88%, an industry source told Calibre Defence.<\/p>\n
The exercise included electronic warfare and jamming designed to simulate combat with an adversary like Russia. Russia focused a lot of its development effort on electronic warfare (EW) in a bid to counter what it saw as NATO\u2019s key advantages, namely its command-and-control infrastructure, and precision strike weapons. It put its full suite of EW systems to use in Ukraine, leading to a constant cycle of adaptation and counteradaptation between the two forces.<\/p>\n
It is worth noting that Helsing has been working in Ukraine for some time. It initially deployed its HF-1 strike drone to Ukraine, a simpler predecessor to the HX-2 tested by the US Army. Through that exposure, the company has likely gained some experience and knowledge that make the HX-2 resilient to jamming.<\/p>\n
HX-2 as a recce strike platform
\n![\"Helsing](\"https:\/\/www.europesays.com\/defence\/wp-content\/uploads\/2026\/07\/hx-2-asgard-800x450.png\")
In separate tests with the British armed forces, the Helsing HX-2 strike drone has demonstrated its ability to hit various targets. Credit: Helsing<\/p>\n
The US Army\u2019s Chief Technology Officer, Alex Miller, told Axios that soldiers successfully used the HX-2 as both a reconnaissance platform and a loitering munition because it could find and track targets \u201ceven under jamming\u201d.<\/p>\n
This may be related to the AI pilot used by the HX-2, which does not rely on a continuous GPS signal or manual pilot data connection. Instead, it employs an onboard AI, likely built from a set of different algorithms, and cameras that allow the drone to map read itself. It can also re-identify, and track targets completely autonomously once launched and given permission to strike, according to Helsing.<\/p>\n
This hints at one of the use cases developed in Ukraine, which has been observed in Russian units employing the Lancet loitering munition. Most long-range loitering munitions have a relatively short endurance \u2013 40 minutes is quite common. The result of this is that they generally require an additional drone to provide detection and target acquisition for them. The loitering munition is then launched for a subsequent engagement, which introduces delay as the munition makes its way to the target.<\/p>\n
However, if the munition is able to effectively perform its own reconnaissance, then it can engage a target immediately upon detection. However, the risk is that no target will be found and the munition lost.<\/p>\n
Calibre comment: Learning from the US<\/p>\n
The US Army tests build on a string of successful demonstrations for Helsing and the HX-2. In a teaming with ARX Robotics<\/a>, the drone was used as part of a recce-strike system, with target spotting carried out by a Gereon UGV. Multiple tests and demonstrations have been carried out in the UK, and contracts have been awarded by Germany for the Bundeswehr and Ukraine.<\/p>\n The US approach to war development is interesting, with a lot of power and decision-making devolved to the lower echelons. It has led to the rapid introduction of a lot of different technologies and significant expansion of its industry. In theory, the US forces are now better equipped than they were in 2022, at least as far as drones are concerned. There is potentially a lot to learn from this approach for European states; the UK\u2019s Project RAPSTONE<\/a> empowers some devolved procurement, for example. But European experimentation contracts tend to be small, and even a procurement can pale in significance compared to a US experimentation award.<\/p>\n