Who this lands on, and who can stop it
Fourteen roles are served. Ten already exist in any incident organisation; four are created by the programme, and each of those is a position an agency would have to fund, train and staff. That is stated as a cost, not a feature. Seven journeys are specified; all seven are below.
The authority asymmetry
One property of this model is worth more than the rest of it: the roles that can stop an action are not the roles that can command one.
| Role | Can | Cannot |
|---|---|---|
| Agency of jurisdiction | Owns the order over its ground, and blocks one outright where jurisdiction is unresolved | — |
| Fire manager / duty officer | Authorises every release, by name, with an expiry | Cannot override a ground crew’s no-drop hold |
| Air tactical supervisor | Can clear the airspace absolutely | Cannot order a release |
| Ground resources | Hold a veto over any drop near their position, effective immediately, with nobody in the approval path | Cannot order a release, cannot task aircraft |
| Detection triage analyst | Dispositions every claim; promotes a recurring false source to the registry | Cannot dispatch resources or confirm a fire into an incident |
| Flight coordinator | Supervises the fleet in flight — the ratio that is the business case | Cannot authorise a release |
| Uncrewed ops supervisor | Airworthiness state and grounding | No requirements exist for this role at all — a declared gap |
A crew member's no-drop hold beats the duty officer's authorisation, and there is no screen anywhere in the system that resolves it the other way. That is not a courtesy. It is the recognition that the person who can see the trees holds information the person with the authority does not — and that the interface has to encode which of them wins before the disagreement happens rather than during it.
The screens page shows what that looks like in practice: the hold screen states, on its face, that no approval was required.
Seven journeys, and what each exercises
A lightning bust in July
District duty officer — the primary user
Dozens of strikes, an unknown number of holdovers, and a resource picture that changes every few minutes. What the programme changes is not the number of detections — it is that each one arrives with its uncertainty attached and its disposition already reasoned: matched against permitted-burn records, checked against known false sources, and confirmed from a second angle before it becomes an alert.
What reaches this person is a claim they can act on, not a pile of alarms to triage.
A twelve-hour shift in August
Detection triage analyst — a role the programme creates
Roughly thirty claims an hour, each needing a disposition in about two minutes. The design consequence is severe and specific: everything needed to decide sits on one surface, because any layout requiring navigation collapses at that rate.
A new role is a cost, not a feature. An agency has to fund, train and staff it.
Arriving over a growing incident
Air tactical group supervisor
This role can clear the airspace absolutely and cannot order a release. The uncrewed fleet yields reflexively on any indication of a crewed aircraft — descending or landing without waiting to confirm, treating ambiguity as confirmation.
Correct as safety engineering, and simultaneously a cheap, permanent denial of service against the fleet. The corpus records both.
Working a line under programme aircraft
Engine boss and crew
A rugged tablet, oriented to the user’s heading rather than north-up, with two large controls: set a no-drop hold, or request a look over the rise. The hold takes effect immediately and the screen says plainly that no approval was required.
The crew must know the veto is theirs, or they will not reach for it when it matters.
The fire on the wrong side of the line
District duty officer again — the journey the design is written for
The system can see it. It can tell the neighbouring authority automatically. It cannot act on it. The refusal is engineered, explained in operational language, and logged as a permission gap rather than a capability gap.
Those two are never summed. A programme that conflates them will eventually argue for authority it has not earned.
Nine aircraft, one queue
Flight coordinator — the role whose ratio is the business case
Pre-flight at 05:45 with one prioritised queue rather than nine displays. At 06:10 an aircraft refuses to launch — not for a technical fault, but because its planned leg has no reachable pre-surveyed emergency landing area since one site’s survey expired. At 11:00 wind rises, the relay descends, the coverage volume contracts, and two aircraft outside the new volume are recalled before they lose link. At 15:00 this role refuses a drop the duty officer had authorised, because an aircraft was transiting the release corridor — and does not ask permission to refuse.
The 06:10 refusal is the unglamorous one that matters: an expired survey, not a fault. It is the check that keeps the lost-link terminal action from being fiction — a planned descent to a site nobody has surveyed is not a plan.
The season, and the inquiry
Agency administrator
After an incident, the question is whether the organisation can reconstruct what happened — including the interface state each human was shown, and refusals at the same fidelity as actions — without the supplier in the room.
An investigation you cannot run yourself is one you do not control, and it cannot be retrofitted: the uncaptured data is not recoverable.
Two asymmetries the journeys keep exposing
- Telling across a boundary can be automatic; acting across one never is. Notification crosses jurisdictional lines freely. Authority does not, and the interface removes the control rather than refusing the click.
- A permission gap and a capability gap are never summed. Both are logged, typed and counted separately, so that a refusal which would otherwise vanish becomes next year's negotiating position rather than an argument for widening authority now.
None of these people has seen any of this. No agency has been approached, no duty officer has reviewed a screen, and no crew has held one of these tablets. The journeys are reasoned from published role definitions and incident-command practice, not from fieldwork — and the people who would know whether they are right are exactly the ones who have not been asked. The contact page reaches the author directly.