Whilst we will always support and welcome the development of Uncrewed Systems and swarming capabilities, we think there are some gaps in the work being done. Being able to physically field the swarm is as important as developing the technical capability.
Jack Tawney - Uncrewed Systems Lead, Frazer-Nash
Drone Swarms: not all about the software.
Recent years have seen an acceleration in research on autonomous systems and the development of drone swarming concepts. Namely, advancing autonomy algorithms, enabling cooperative behaviours, ensuring safety and developing software interoperability standards.
But this growing body of work has obscured a more fundamental challenge. While progress has been made in designing how swarms think and behave, far less attention has been paid to how they are to be fielded at scale under operational conditions.
The result is a widening gap between hypothetical swarm management systems and practical capability. In short, the discipline of fielding the swarm remains underdeveloped, so we risk developing fleet management systems without fleets to manage.
Deployment and ‘fielding’ the swarm remains underdeveloped.
Last month marks one year since Operation Spider’s Web, a coordinated drone strike employing more than 100 systems that shocked the world in its audacity and impact on Russia’s aviation capability. While notable for its scale and effect, the operation did not depend on novel autonomy.
Rather, it was enabled by innovations in deployment: covert transport, concealment, and coordinated launch of relatively simple systems brought into theatre in a ready-to-use state. The decisive advantage lay not in swarm behaviour once airborne, but in how systems were prepared, positioned, and rapidly launched.
By contrast, the West continues to experiment at modest scale, often with a dozen or fewer systems under controlled conditions. This risks assuming that scaling from tens to hundreds or thousands is largely linear; true for some software sub-systems, but not for most other development lines.
Transitioning from demonstration to operational capability introduces different challenges: storage, transport, power management, and integration with existing platforms and C2. Riskier still is the use of simulation to generate “quasi mass”, which, while economical, abstracts many of the real challenges of multi-UAS operations at scale.
These considerations are not merely procurement issues. The safe handling, platform integration, and rapid deployment of large numbers of drones, some configured as munitions, potentially in contested environments and under time pressure, demands innovation no less sophisticated than autonomy.
Yet research programmes have prioritised cooperative behaviours, assurance mechanisms, and software-focused interoperability standards. While essential, they have not prepared forces to absorb and operate swarms at meaningful scale, nor significantly advanced the “science” of sustaining high-readiness swarm deployments.
Moving swarms from concept to practical reality.
If swarming is to move from concept to capability, experimentation must become more persistent, integrated with operational units, and focused on end-to-end deployment. The task is not to reduce investment in autonomy, but to rebalance effort towards fielding, deployment, and integration.
That rebalancing requires practical interventions. Here’s our ideas of what they could be:
1. A concerted push into mass deployment.
Focus on stowage, transport, and launch concepts capable of bringing hundreds or thousands of drones into theatre in a ready state. This could include a combination of retrofitting existing aircraft, vehicles and vessels with modular launcher systems, as well as exploring how existing civilian vehicles and static infrastructure can be adapted. What matters is the ability to generate mass, maintain readiness, and deliver reliable launch at scale. Without this, advances in autonomy risk outpacing the means to employ them.
2. Routine experimentation.
Swarming trials remain bounded by programme lifecycles and showcase events, reinforcing the perception of swarming as exceptional and novel. Instead, experimentation should be continuous, frequent, and iterative. A shift toward maintaining ready inventories for routine use via a “drone swarm as a service” model would force the development of professionalised deployment processes while enabling wider access and normalising swarm use. This would surface operational realities faster and allow use within opposing force training.
3. “Project 300”.
A deliberately challenging tri-service exercise to launch 100 drones each from in-service land, sea, and air platforms simultaneously. This would make mass deployment tangible, expose integration and regulatory challenges, and demonstrate force multiplication as a strategic signal with deterrent value.
The trick is to bring all the elements together
We at Frazer-Nash have had the privilege of supporting the swarming R&D programme in the UK for many years now, working very closely with both SMEs and primes alike, and our team come from a diverse background that includes ex-MOD civil servants, analysts and scientists, reservists and veterans from all services.
We’ve a long track-record of delivering proofs of concepts and achieving some UK / US “firsts” in uncrewed system capabilities…but increasingly we’re branching out into mission-partnering and managed service provision models, where we co-create the really big ideas with public sector clients, and deliver on the mission objectives in a solution and supplier-agnostic way, at pace.
Want to co-create “Project 300” with us?
What have we missed?
Drop a line to Jack Tawney j.tawney@fnc.co.uk
-
Share:
-
-
-
-
-
