Tasking many as one
How a single operator directs a group without workload growing with its size.
Research · Swarm Autonomy
Swarm
autonomy.
A swarm is not simply a lot of drones. It is a group of autonomous systems that behave as one team — dividing a task between them, sharing what each of them sees, and continuing to work when individual members drop out. That last property is what makes the idea powerful.
The Idea
Resilience through
numbers.
A single capable platform is a single point of failure. Lose it and the mission stops. A group of cheaper, individually less capable platforms fails differently: losing one degrades the group slightly rather than ending the task, and the cost of that loss is proportionate.
That changes what is reasonable to attempt. Tasks that would be unacceptable to risk an expensive aircraft on become straightforward when the platform involved is inexpensive and the group can absorb its loss. It also changes coverage — many sensors in many places see more than one sensor in one place.
What Makes It A Swarm
Coordination,
not just quantity.
The engineering question is how a group reaches sensible collective behaviour without a central controller telling every member what to do — because that controller becomes the single point of failure the swarm was meant to eliminate.
- Task divisionSplitting an objective between members and reallocating when circumstances change.
- Shared pictureCombining what each member observes into an understanding none of them had alone.
- Decentralised decisionsLocal decisions that add up to sensible group behaviour without constant central direction.
- Graceful attritionContinuing to function, with proportionate loss of capability, as members drop out.
SwarmDecentralised
Unit of valueThe group, not the individual
Failure modeDegrade, not stop
CommunicationAssumed limited and intermittent
CommandOne operator tasks the group
StageConcept-level research
Open Questions
What we're
still working out.
This is early research. We describe the shape of the problem rather than claiming it solved.
Talking sparingly
Coordination that survives very limited and unreliable communication between members.
Avoiding each other
Many platforms in a small volume without collisions or interference.
Predictable group behaviour
Emergent behaviour that a commander can actually anticipate.
Cost per unit
The economics only work if individual members are genuinely inexpensive.
Accountability at scale
Keeping human command meaningful when the group is large.
Swarm autonomy is a concept-level research direction at Rewind Dynamics, described here at a high level by design. Everything described here is under active research and development. Nothing has been fielded, and we do not publish specifications or performance figures. Our first end-to-end mission test is planned for December 2026. Technical detail is shared directly with partners, evaluators and the armed forces under appropriate terms — get in touch.
Go Deeper
Interested in
multi-agent autonomy?
If you work on decentralised coordination or multi-agent systems, we'd like to talk.