The DECIDE Model
PAVE tells you where to look for hazards. DECIDE tells you what to do once you've found one — whether that's a marginal weather call on the ground before launch, or a genuine surprise mid-flight, like an unexpected loss of GPS lock over a crowded worksite. DECIDE is a six-step process, one letter per step, built to work under real time pressure rather than as a leisurely planning exercise. This lesson walks through all six steps and applies them to both a preflight decision and an in-flight one.
D — Detect that a change has occurred
The process starts with noticing something has changed from what you expected or planned — wind picking up beyond the forecast, an unexpected person walking into the operating area, a battery voltage reading lower than it should be at this point in the flight, or a control link warning on the controller's display. Detection sounds obvious, but it's the step most often skipped by a pilot who is task-saturated, distracted, or simply not scanning for changes because the flight has been uneventful so far.
E — Estimate the need to react
Once a change is detected, the pilot estimates how significant it actually is and how urgently it needs a response. Not every detected change requires immediate action — a slight, momentary wind gust within the aircraft's rated tolerance may need only continued monitoring, while a rapidly dropping battery voltage or a person entering the flight path needs an immediate response. This step is where a pilot's judgment about severity and urgency actually gets exercised, rather than either overreacting to something trivial or underreacting to something serious.
C — Choose a desirable outcome
Before jumping to a specific action, the pilot chooses what a good outcome actually looks like for this situation — land immediately at the nearest safe point, return to the original launch point, continue the mission but adjust the flight path, or abort and reschedule the whole operation. Naming the desired outcome first, before choosing a specific action, keeps the pilot focused on the actual goal (a safe resolution) instead of jumping straight to the first action that comes to mind.
I — Identify actions to control the change
With a desired outcome chosen, the pilot identifies the specific, concrete actions available to achieve it. If the chosen outcome is "land immediately at the nearest safe point," the identified actions might include confirming the nearest safe landing zone is clear of people, switching to a manual control mode if automated return-to-home isn't appropriate for the situation, and communicating the plan to a visual observer or ground crew member if one is present.
The six-step DECIDE model as a continuous loop — evaluation feeds back into detection for as long as the situation continues to evolve.
View LargerD — Do the necessary action
This is the step where the pilot actually executes the identified action — not just plans it. A common failure mode is stopping at "identify" and hesitating on execution, especially when the action feels drastic (aborting a paid mission, declaring a return-to-home) or socially awkward (telling a client or crew member the flight is stopping). DECIDE only works if the pilot commits to acting once the prior steps are complete.
E — Evaluate the effect of the action
After acting, the pilot evaluates whether the action actually produced the intended outcome. If the aircraft returned to home safely and the situation is resolved, the loop is complete. If the situation has changed again — wind is still building, the battery is still dropping faster than expected — the DECIDE loop restarts from Detect with this new information. DECIDE is explicitly a loop, not a one-time linear process, precisely because real situations often require more than one cycle to fully resolve.
Worked scenario: DECIDE applied on the ground before launch
You're preparing to launch a real estate photography mission and notice the wind has picked up more than the forecast predicted (Detect). You estimate this is significant enough to matter, since it's now close to your aircraft's rated wind limit (Estimate). You choose a desired outcome: only fly if you can maintain a safe margin below the rated limit throughout the shoot (Choose). You identify your available actions: check a live wind reading at the site, wait 15 minutes to see if the gusts are a passing trend, or reschedule entirely (Identify). You check the live reading, confirm it's trending toward the limit rather than away from it, and reschedule (Do). Afterward, you evaluate that rescheduling was the correct call once you see the wind continued to build through the original time slot (Evaluate).
Worked scenario: DECIDE applied in flight
Mid-mission, you notice the aircraft's telemetry shows a sudden GPS signal degradation warning over a dense urban worksite (Detect). You estimate this needs an immediate response, since degraded GPS affects position hold and return-to-home accuracy (Estimate). You choose the desired outcome: land the aircraft safely under manual control before GPS loss becomes total (Choose). You identify the nearest clear landing area within visual line of sight and the manual flight mode needed to reach it (Identify). You switch to manual control and land at that clear area (Do). Once down, you evaluate that the landing was successful and safe, and that the mission should not resume until GPS reliability at that location is better understood (Evaluate).
| Step | What it means | Common failure at this step |
|---|---|---|
| Detect | Notice that something has changed from the expected plan or condition | Task saturation or distraction causes the change to go unnoticed |
| Estimate | Judge how significant and urgent the change actually is | Overreacting to a trivial change or underreacting to a serious one |
| Choose | Select the desired outcome before picking a specific action | Jumping straight to an action without a clear goal in mind |
| Identify | List the specific actions that could achieve the chosen outcome | Considering only one option instead of the full range available |
| Do | Actually execute the identified action | Hesitating to act because the action feels drastic or awkward |
| Evaluate | Confirm the action produced the intended outcome, or restart the loop | Assuming the situation is resolved without confirming it |
Applying DECIDE to an In-Flight Surprise
Use this sequence any time something unexpected happens during a mission, however minor it first appears.
Actively scan for changes throughout the flight
Don't wait for a change to demand your attention — deliberately monitor telemetry, the surrounding environment, and the aircraft's behavior throughout, not just at launch and landing.
Judge severity and urgency honestly
Resist the pull to downplay a change just because acknowledging it is inconvenient mid-mission.
State the desired outcome out loud
Even flying solo, explicitly naming the goal ("land safely at the nearest clear point") focuses the next two steps and prevents jumping straight to an unconsidered action.
List more than one possible action
Even under time pressure, briefly consider at least two ways to reach the desired outcome before committing to one.
Commit and act without further delay
Once the action is identified, execute it — hesitation at this stage erodes whatever safety margin the earlier steps preserved.
Confirm the outcome, then loop again if needed
Verify the action actually resolved the situation. If it didn't, or a new change appears, restart the DECIDE loop from Detect rather than assuming the problem is over.
| Mistake | Why it happens | Correct understanding | Regulation / source |
|---|---|---|---|
| Treating DECIDE as only for in-flight emergencies | The step names sound reactive and urgent | DECIDE applies equally to a calm, preflight go/no-go decision as it does to an in-flight surprise. | ACS Task V.D |
| Stopping after identifying an action without actually doing it | The action feels drastic or uncomfortable to carry out | The "Do" step requires actual execution; identifying an action without acting on it provides no safety benefit. | FAA Remote Pilot Study Guide |
| Treating DECIDE as a strictly one-time, linear process | The six letters read like a simple sequential list | DECIDE is a loop; the Evaluate step can restart the process from Detect if the situation isn't fully resolved. | FAA Remote Pilot Study Guide |
| Skipping the "Choose" step and jumping straight to an action | Time pressure makes naming a goal feel like a wasted step | Choosing the desired outcome first keeps the identified action aligned with an actual safety goal, not just the first idea that comes to mind. | FAA Remote Pilot Study Guide |
Does DECIDE have to be applied out loud or in writing?
No. In practice, an experienced pilot runs the six steps mentally in seconds for a minor situation. Verbalizing it (even alone) is a useful training habit while the model is still new, and remains valuable when a visual observer or crew member is present to hear the reasoning.
Is DECIDE required by a specific Part 107 regulation?
No regulation names DECIDE directly. It is FAA risk-management guidance tested under ACS Task V.D, supporting the broader responsibility 14 CFR 107.19 places on the remote pilot in command to determine the aircraft is safe to operate throughout the flight.
What happens if I detect a change but decide no action is needed?
That is a valid outcome of the Estimate step — not every detected change requires a response. The important part is that the judgment was made deliberately (estimating the change as low severity and low urgency), not simply ignored.
Can DECIDE be used for decisions that have nothing to do with an in-flight emergency?
Yes. DECIDE works for any structured decision under uncertainty, including administrative and planning decisions like whether to accept a job with an unusual scope, or whether newly purchased equipment is ready to use on a paid mission.
Test Your Knowledge
Answer the questions below to check your understanding. Every answer can be found in the lesson above.