Practice. Decide. Improve.

Why this exists

The Pilot Method combines current source editions, airport-specific scenarios, your own performance history and optional instructor review. Static information is not the moat — the loop is.

Generative models may interpret language or render a voice. Deterministic scenario rules and reviewed source data decide anything safety-critical. Providers stay replaceable.

Real today

• — Twelve distinct FXE development-preview scenarios run end to end: closed traffic to a full stop, closed traffic for the option with a simulated go-around, the local practice area with flight following, the requested shoreline transition and return, a runway change detected before taxi, a controller readback correction, progressive taxi after losing the picture on the surface, the return arrival with a traffic sequence and runway exit, a land-and-hold-short offer you may decline, a blocked transmission followed by a say-again, a suspected radio failure resolved with light signals, and a traffic advisory answered honestly.
• — ATIS turns are copying exercises, not multiple choice: you fill a kneeboard-style scratchpad — code, wind, visibility, sky, altimeter, runways and advisories — and each field is scored on meaning. No value appears in any label or placeholder before you submit.
• — Values you say that were never issued — another facility's frequency, a runway you were not assigned, a guessed hold-short — are flagged as conflicting elements and block the turn the same way an omission does.
• — The session carries state forward: ATIS code, runway, wind, altimeter, beacon code, altitude, the facility you are working and any outstanding instruction stay visible while you fly the scenario, and the debrief lists the exact FAA sources and paragraphs read for that scenario.
• — Audio first: every controller transmission is played, not read. The transcript stays hidden until you answer, and revealing it early is recorded as an assisted turn in the debrief.
• — Clearance Delivery is a first-class phase, exercised in the scenarios where this synthetic case includes a VFR clearance and explicitly noted as not required in the ones where it does not.
• — Deterministic evaluation accepts reordered and abbreviated phrasing, blocks safety-critical omissions, and finishes with a full turn-by-turn debrief including strengths and what to practise next.
• — Reviewed source cycles and effective dates are stored as data. The lab selects the controlling cycle by date and closes airport-specific scenarios once the last reviewed cycle ends.
• — Signed in, your own transcripts and scores are saved privately to your account. Audio is never recorded or stored.

Not available yet

• — Server-side aviation transcription is implemented and fail-closed: the endpoint exists, sends only the clip you push to talk, stores nothing, and stays disabled until the server-only OPENAI_API_KEY secret is added by the owner. Until then the input is labelled “Browser transcription — review before grading.” and you edit the text before it is graded.
• — Premium realtime voice is not configured. Controller audio in the preview is synthetic browser speech generated on your device, clearly labelled and never a real controller.
• — Standard and Busy difficulty are locked; only Guided is implemented.
• — Scenario content is reviewed against the FAA sources named in each debrief. That review is not a line-by-line operational sign-off and not an FAA approval.
• — ATC Lab is free while it is in development. No pricing, checkout or paid access is open for it.

Free development preview. No price exists and nothing is for sale. Create an account to be considered for access. Creating an account does not grant access and does not charge anything.