A formation with no pilot on board
Piloted formations adapt because pilots talk to each other. A UAV squad has no one on board — the case for shared state over a secured link.
SquadSense is coming soon. This article sets out the problem SquadSense is designed to solve and the approach it takes, so that vehicle makers can plan for it.
The gap: isolated aircraft
Manned-aircraft formations have pilots who coordinate and adapt in flight. A piloted formation adapts because the pilots talk to each other: when one aircraft has a problem, the others know and adjust.
A UAV squad has no one on board. In contested airspace, its ground link may be jammed or degraded at the very moment something goes wrong. UAVs that fly as isolated aircraft cannot react together when a companion loses navigation, link or system health. An aircraft that cannot see its companions cannot react when one of them fails.
An aircraft that cannot see its companions cannot react when one of them fails.
The response: shared state
The fix is not to put a pilot back on board. It is to give each aircraft what a pilot would have had: awareness of the others. If aircraft share their state, the squad can make bounded adjustments together as conditions change.
"Bounded" is the important word. The aim is not open-ended autonomy, but a limited, predictable set of adjustments — retasking around a failure within limits that the operator and host system have set.
What SquadSense will do
SquadSense, coming soon, is the squad layer in ShoalOS, sitting at L3 in the eight-layer stack, directly above the onboard navigation services. What it will do:
- Sense the nearby squad — each payload shares its navigation confidence and payload health with the rest of the squad.
- Secured data link — state travels between squad members over a secured link.
- Retask — when a sensor, drone or mission objective fails, the team can retask the aircraft it has available.
- SquadSense SDK — shares position quality, link status and system health across the team.
The data comes from SenzFusion. Because every payload already maintains its own navigation estimate — built from IMU prediction, radar map matching and a thermal scene check, with GNSS used as a check for contested environments — it can report how confident it is in that estimate and how healthy its sensors are. SquadSense, coming soon, will share exactly that.
The idea is shown in the Stadium Race concept film ↗. It is a concept film, not a demonstration of a working system.
People stay in the loop
Key decisions stay with the operator and the host. The planner / operator view shows team state and reassigns tasks; the squad's job is to keep that view accurate and to make the bounded adjustments it has been allowed to make.
For a vehicle maker, the practical point is that SquadSense builds on the same payload and SDK used for navigation. Integrating SenzFusion today is the foundation for squad coordination later.
Read more about SquadSense in the Systems overview, or see L3 on the Platform page.
