Counter-UAS and short-range air defence

Low, slow and small air threats defeat the tempo of conventional air-defence procedure. A crew may face dozens of simultaneous detections, most of them clutter, birds, friendly systems or the same object seen by three sensors at once. The time between a credible detection and the last moment an effect can be applied is often measured in seconds.

In that window the crew must de-duplicate tracks, classify the object, confirm it is not friendly or civilian, check the airspace control order and the rules of engagement, choose a proportionate effector, and act. Doing that reliably under pressure, and being able to show afterwards that it was done correctly, is the problem.

The profile fuses the sensors a SHORAD or counter-UAS site already fields:

ISR Fusion correlates detections into a single track, resolving the common case where radar, RF and EO/IR each report what is actually one object, and attaching a per-sensor classification confidence rather than a single opaque score.

The threat-assessment function of SC-ISR-01 ISR Fusion Agent evaluates the track against protected assets, its behaviour and its likely payload class. The deterministic policy-enforcement service evaluates the airspace control order, weapon control status and rules of engagement deterministically, as code rather than as model output, and marks each candidate response as permitted, prohibited or requiring higher authority. The effects-planning function of SC-TGT-01 Targeting & Engagement Agent then proposes a proportionate effector, considering collateral risk, fratricide risk, magazine state and the geometry of the engagement.

The whole chain is designed to complete inside the crew's decision window and to log every step as it goes, not afterwards.

The crew sees one track list, ordered by urgency, with the reason for each ranking visible. Selecting a track shows the contributing sensors, the classification and its confidence, the applicable rules and the recommended response with its alternatives.

Interaction is deliberately spare at this tempo: confirm, choose an alternative, hold, or reject. Free-form querying is available before and after an engagement rather than during it, and the crew can adjust filters and thresholds between engagements, with each change recorded.

SynapseCommand correlates the threat picture, recommends a response and prepares the evidence required for an authority decision. It does not fire or command an effector directly. Following explicit operator approval, the authorised recommendation is passed to the existing host fire-control system, which remains responsible for effector tasking and execution.

The sequence is explicit: threat track correlated, recommended response generated, human-in-the-loop authority requested, operator approval received, authority record verified in the hash-chained decision log with ML-DSA signing in assured deployments, approved recommendation passed to the host fire-control system, host fire-control system tasks the authorised RF effector, and effect assessment returned to SynapseCommand. There is no autonomous engagement mode in this profile.

Where a candidate response is prohibited by the rules of engagement, the system says so and does not offer it. Where it requires higher authority, the request is routed with the evidence attached. Every authorisation is written to the hash-chained, append-only decision log, ML-DSA-signed in assured deployments, and bound to the track state, the applicable rule set and the time, so a post-incident review can reconstruct exactly what the operator was shown.

Each engagement or non-engagement produces:

The platform sits between the sensor layer and the existing air-defence C2, publishing its fused track picture and recommendations into the crew's current display where the interface permits, and running as a parallel decision-support console where it does not. Effector interfaces remain under the control of the fire-control system; SynapseCommand recommends into it and never commands it directly.

Deployed at the tactical edge on ruggedised hardware, operating fully air-gapped where required. The site continues to function with no connectivity to any wider network; policy sets and model versions are updated through a controlled, signed offline process, and every update is recorded in the same evidence chain as operational decisions.

The benefit is a shorter and more consistent decision cycle. Fusing detections removes the duplicate-track workload that dominates crew attention, and pre-evaluating the rules of engagement removes the step most likely to be rushed. Because the evidence is captured as the engagement happens, after-action and legal review draw on a complete record instead of recollection.

Figures quoted anywhere on this site are illustrative planning targets from internal test scenarios, not certified operational results.

Classification of small air systems is genuinely hard: sensor coverage is uneven, adversary systems change quickly, and civilian and military airframes overlap. The platform reports confidence honestly, including low confidence, rather than presenting a clean answer it cannot support.

It cannot compensate for sensors that do not detect the object, for a stale airspace control order, or for rules of engagement that have not been encoded. Encoding those rules correctly is a joint task with the customer's legal and operational staff, and the system's behaviour is only as sound as that encoding.