Closing the Loop: Enhancing U.S. Drone Capabilities through Real-World Testing

The war in Ukraine has been a wake-up call for global militaries, demonstrating the critical role of small and cheap commercial drones. The Department of Defense (DOD) acknowledges the strategic importance of attritable, all-domain unmanned aircraft systems (UASs), recognizing that leveraging commercial technology is crucial for staying competitive in the innovation race. But U.S. commercial drones sent to Ukraine have faced significant challenges on the battlefield, which poses one critical question for the DOD: How can it ensure U.S. drone manufacturers can adapt quickly enough to meet changing battlefield demands and maintain the technological advantage?

The DOD’s response to close the gap in expendable drone capability has mostly been limited to acquisition efforts, which are only one part of the challenge to keep abreast of U.S. competitors in the drone arms race. So far, the Defense Innovation Unit (DIU) has created a Blue UAS program, containing policy compliant commercial unmanned aircraft vehicles (UAVs). These UAVs are validated as cybersecure and safe to fly and are available for government purchase and operation. The DOD also has an Army program of record for small quadcopter drones—short-range reconnaissance small unmanned aircraft systems—with a FY 2025 budget request of $21.1 million to acquire 540 systems, with an expected cost per system of $65,000. The most prominent initiative is the $1 billion Replicator program for buying “small, smart, cheap, and many” all-domain unmanned aircraft systems with 30 contracts for hardware and software projects already awarded, involving over 50 major subcontracts. Since 2023, the United States Special Operations Command’s small unmanned systems program focuses on acquiring and developing solutions for special operations using off-the-shelf technology with a budget of $14.6 million in FY 2024. This program involves developing and testing AI/ML to enhance drone autonomy, including autonomous navigation, obstacle avoidance, automated target recognition, and multi-drone operations controlled by a single user.

While these efforts demonstrate the DOD’s intention to reduce barriers and increase funding for the entry and acquisition of commercial technology, simply buying drones is not enough. The DOD should ensure that these drones are effectively utilized by military forces, relevant to mission tasks, and technologically superior to those of adversaries. A critical aspect for achieving these goals is the software that powers unmanned systems, which relies on advanced algorithms and neural networks that must be trained on real-world data to perform optimally. In conversations with CSIS, Ukrainian military UAS operators noted that it often takes less than a week for both sides to find ways to trick algorithms and AI models. For instance, a simple tactic like placing tires on top of a tank can disrupt computer vision and confuse the system.

It is crucial for the DOD to complement its improved acquisition process by incorporating Ukraine’s experience in enhancing adaptability, refining testing environments, improving training, advancing AI deployment, and—most importantly—establishing effective feedback loops on drone performance in the battlefield. Achieving this is impossible through testing in safe environments, like Arizona test ranges, which are vastly different from actual war zones. As representatives of the Armed Forces of Ukraine (AFU) shared with CSIS, U.S. drones with autonomous features that had been recently tested in the United States were quickly lost when flying over the forests of Kharkiv. The computer vision systems struggled to identify objects amid the dense “ocean” of green and navigate, especially when GPS signal jamming further exacerbated the problem.

Some might argue that the drones the DOD needs for a potential conflict with China differ from those used in Ukraine and that U.S. drones’ success in Ukraine is less important. However, this view is mistaken. It’s not just about testing performance or technology in a specific environment; it’s also about understanding the creativity of the adversary. The battlefield reveals the enemy’s technological advantages and know-how, which are crucial factors that must be accounted for, regardless of the region or opponent.

Fully understanding the importance of the adversary’s perspective and the potential of U.S. commercial drones, drone companies like Skydio, AeroVironment, Red Cat, Freefly Systems, and PDW rushed to test their drones on the front line by donating small quantities and making brief trips to Ukraine. In conversations with CSIS, Ukrainian military officials noted that these companies struggle to maintain effective communication between frontline operators and the manufacturers. The fast-paced evolution of drone and counter-drone tactics means that manufacturers who drop off their products and then disengage are of little use. Ukrainian companies are successful in competing against Russia not necessarily because they started with superior technology, but because they have rapidly evolved and continuously made changes in response to direct feedback from the front lines.

The Ukrainian experience with rapid iterations and improvements could benefit U.S. drone manufacturers and provide a model for their involvement. By examining Ukrainian practices, U.S. military organizations and companies can identify ways to enhance the feedback loop and improve existing processes. Three key aspects of the Ukrainian experience should be considered for deeper U.S. engagement:

1. Fast and Easy Testing: Ukraine develops testing ranges with simplified access procedures and professional evaluations, as well as facilities for assessing robotic systems, drones, and electronic warfare systems. One of the examples is the recently launched Iron Range testing facility, a combined effort of the General Staff of the Armed Forces of Ukraine and the Ministry of Strategic Industries. “The facility aims to provide quick feedback during the design phase, helping manufacturers save time and speed up the deployment of essential weapons and equipment to the front lines,” the deputy minister of strategic industries Anna Gvozdiar. CSIS conversations with Ukrainian drone manufacturers often mentioned the so-called Donetsk Zone, a location designed to simulate real combat conditions. The area features artillery craters, hills, and flat terrain, replicating the challenges that military equipment, especially electronic warfare systems, would encounter in battle.
   
   In the United States, there is extensive testing infrastructure, such as the White Sands Missile Range and the Nellis Air Force Base. However, as Chris Bonzagni, a former DIU program manager and founding member of the Blue UAS initiative, told CSIS in the interview with research team, U.S. drone producers face significant challenges accessing testing grounds, especially for electronic warfare capabilities. Dual-use drone manufacturers, which the DOD aims to engage through its new commercial drone acquisition programs, struggle with testing scheduling, often facing months of setbacks and prohibitive costs. Long testing time means delays, as it can take up to several months to return for further tests, significantly slowing down development cycles.
    
2. Frontline-Localized R&D: Ukrainian drone manufacturers and brigade commanders shared with CSIS that many military units have set up workshops directly on the front line to modify and repair drones, streamlining research and development (R&D). Often located right in the trenches, these small tech labs allow drone producers to bring their products directly to the battlefield. This enables engineers to see how the drones perform in real time, get immediate feedback, and make quick adjustments, including software updates, without needing to send the drones back from the front line.
   
   Both manufacturers and warfighters admitted in conversations with CSIS that this hands-on collaboration is highly effective as it balances engineering and military approaches to innovation. Engineers might add unnecessary features, use AI where it’s not needed, or miss opportunities to apply it efficiently, which can increase costs and complexity of the end product. On the other hand, military personnel may push for overly complex solutions, being used to older, larger systems. Direct collaboration ensures a practical and cost-effective approach, focusing on real battlefield needs, which aligns with the U.S. goal of producing drones at a lower cost without over-specifying them.
    
3. Local Representation and R&D: Ukrainian drone companies benefit from close communication with the military, having their main R&D facilities in safe cities. This setup allows military personnel to visit them and directly convey needs to engineers, who can then quickly develop and test solutions on the ranges or send them to the front line. “This rapid, task-specific process, combined with ongoing training and feedback, significantly improves drone performance, boosting target accuracy from 10% to 70-80% within a single upgrade cycle,” says founder of a drone manufacturer Vyriy Drone Oleksiy Babenko.

The testing and frontline conditions in Ukraine are unparalleled and cannot be fully replicated in the safe environments of nonwar countries, including the United States. As a result, U.S. drone manufacturers and the military should find ways to engage in Ukraine and tap into Ukrainian expertise. There are two ways to do that.

First, U.S. drone companies could set up local offices in Ukraine with both U.S. and Ukrainian engineers. Ukrainian engineers would translate military needs into specifications and provide ongoing support, while U.S. engineers could ensure seamless production back home. However, U.S. early-stage start-ups may struggle with the costs, International Traffic in Arms Regulations, and other restrictions limiting U.S. citizens from working with dual-use or military technology in Ukraine.

A few companies are already taking these steps, which is a positive trend for the U.S. commercial drone sector. Skydio is hiring a small team in Ukraine, focusing on engineering and customer support. Its CEO Adam Bry noted that Ukraine’s sophisticated drone users are now driving the company’s product development, even if it doesn’t overlap with U.S. government requirements. Another example is Wilder Systems, which is seeking partners in Ukraine for its low-cost drone program and testing them on the battlefield. However, these examples are exceptions rather than the norm. More companies should follow suit, and the U.S. government should actively find ways to encourage this involvement.

Second, to expand testing opportunities for U.S. start-ups and gather battlefield feedback, a hybrid collaboration model with Ukraine could be effective. When the AFU identifies a need, the DIU or private organizations could quickly source candidates from the U.S. drone market. These drones would undergo short-term testing in Ukraine under battlefield-like conditions, with AFU representatives providing feedback. The top-performing drones would then be deployed with operational units for one to three months on the front line, with a local product manager facilitating real-time feedback, rapid development, and software updates. After evaluation, the technologies that perform the best could be issued U.S. contracts for delivery to Ukraine and be considered for U.S. programs such as Replicator, Blue UAS, and others. This approach would not only accelerate innovation but also provide Ukraine with more relevant drones tailored to its needs and in the larger quantities required for the war.

The true battleground for drone supremacy lies not only in procurement pipelines or controlled test ranges, but in the unpredictable chaos of real-world conflict. Ukraine’s war-torn landscape has become a crucible where technological ingenuity and rapid adaptation determine survival. For the United States, engaging in this crucible is not just a strategic choice—it is a necessity. The United States should seize this unparalleled opportunity, embedding its drones on the front lines to enhance capabilities and build an adaptable, evolving drone force ready to confront the unexpected challenges of future conflicts.

Kateryna Bondar is a fellow with the Wadhwani Center for AI and Advanced Technologies at the Center for Strategic and International Studies (CSIS) in Washington, D.C.