We rolled eye-tracking into our AATD last week (Varjo XR‑4 with gaze heatmaps), and across three 40‑minute IFR sessions students cut dwell on the attitude indicator by about 25% and held altitude to ±80 ft in moderate chop. It finally lets us measure scan discipline and train it deliberately — anyone here feeding this data into CBTA/EBT scoring or pairing it with real‑time weather in the sim?
I’m feeding gaze into EBT under “monitoring and cross‑checking” by phase: if AI dwell >6s or time‑to‑airspeed >2s after a speed call, it flags and I pull a 10‑sec clip for debrief. Small caveat: XR‑4 eye calibration drifts after a headset bump or sweaty session, so we run a 30‑sec recal every 15–20 minutes — like tightening your seatbelt after leaning forward.
We’ve had good results starting debriefs with a 20‑sec eyes‑only replay — panel dimmed, just the gaze trace over the PFD — so they call out their own scan gaps; we also suppress EBT flags during COM/QRH windows so your AI‑dwell trigger, @a_knight25, doesn’t false‑hit.
On our AATD with the XR‑4, a soft tone when eyes sit on the FMS/map past about 3 s made debriefs click; students could name the rabbit hole and altitude tightened from ±80 to about ±60 by the second 40‑min block. We tag it in CBTA as monitoring in climb/approach and workload management in cruise, and a quick ‘last look speed/alt’ call helps. Caveat: recal every 20–25 min or the heatmap drifts with glasses, @a_knight25.