Flight - UA972 Chicago → Brussels (14–15 May 2026, Boeing 787-10, Cabin + Cargo Hold)

Griffin Air uses two independent sensors to determine whether the device is airborne. Either the accelerometer or the barometric pressure sensor can initiate flight detection. When the accelerometer detects sustained movement above its configured threshold consistent with aircraft taxi and takeoff motion, it wakes the Ground Service and begins the evaluation sequence. The barometer then confirms whether cabin pressure has dropped more than 25 hPa below the stored ground reference. If both conditions are met and a GPS speed check fails to return a low-speed fix, the device commits to an exited ground mode: all cellular transmissions cease, GPS positional logging is suspended, and sensor data is written to flash memory. This dual-trigger architecture ensures alignment with flight detection system requirements from the aviation authorities. 

UA972 Field Validation - Cabin and Cargo Hold

Two Griffin Air devices were deployed on United Airlines flight UA972 from Chicago O'Hare to Brussels on 14–15 May 2026, operating on a Boeing 787-10 Dreamliner. Device 1396412 was placed in the passenger cabin and device 1396363 in the cargo hold.

Both devices exited Ground Mode at 23:32 UTC, three minutes after the recorded wheels-off time of 23:29 UTC, demonstrating consistent flight detection across both locations.

Throughout the flight, barometric pressure readings remained closely aligned. At cruise, the cargo hold recorded 813 hPa compared with 812 hPa in the passenger cabin, a difference of just 1 hPa, demonstrating similar pressurisation conditions.

Three recorded state checks at 02:05, 03:36 and 05:07 UTC confirmed that both devices remained outside Ground Mode during cruise. At 06:21 UTC, both devices recorded a barometric landing trigger as pressure began recovering during descent.

Post-Landing

The barometric landing trigger at 06:21 UTC indicated recovering pressure but did not automatically restore Ground Mode. Both devices continued evaluating their operating conditions through scheduled Ground Service checks, which occur every 30 minutes by default.

The cargo hold device recorded Ground Mode restoration at 06:51 UTC, just 30 minutes after the barometric trigger. The cabin device restored Ground Mode at 09:33 UTC, 3 hours and 12 minutes after the same event, resulting in a 2-hour and 42-minute difference between the devices.

This difference may be attributed to their respective movement patterns. The cargo device may have satisfied the inactivity conditions sooner, while continued movement of the cabin device during passenger handling and onward transport may have delayed its return to Ground Mode. A successful GPS fix below 100 km/h can activate a temporary five-minute GPS Override, potentially allowing communication without immediately restoring Ground Mode.

These results demonstrate that Ground Mode restoration is not determined by the barometric landing trigger alone. Inactivity, scheduled Ground Service evaluations and any active GPS Override must be considered when interpreting restoration times. The recorded cargo restoration at 06:51 UTC precedes the stated wheels-on time of 07:01 UTC, so that timestamp requires verification before it can be confirmed as full Ground Mode restoration.