Sensor cadence: 60 seconds against 5 minutes

A probe logging every five minutes will catch a four hour excursion. It will describe a fifteen minute one badly, and it will miss the peak by a wide margin when the rate of rise is steep.

On the Changi apron event we reconstructed both versions from the same 60 second data. The full trace peaked at 27.4 °C. Sampled at five minutes, the same event peaked at 26.1 °C and looked half an hour shorter. Neither number changes the release decision here, but the mean kinetic temperature calculation moved by 0.4 °C, and on a tighter profile that gap decides a quarantine.

Battery is the real trade

At 60 second logging with a 15 minute upload cycle, our probes run about 14 months. Drop the logging to five minutes and you get past two years. For a probe that lives in a reusable container and gets serviced annually, 14 months is fine. For a probe glued inside a passive shipper that goes out and comes back three times a year, it is not.

A cadence rule that holds up

• Active containers and anything on a ramp: 60 seconds
• Passive shippers with 96 hour hold time: 5 minutes
• Cryogenic dry shippers: 5 minutes, plus a tilt event trigger that logs at 10 seconds for two minutes
• Road lanes under 12 hours: 60 seconds, because handling dominates

Event driven logging beats a fixed high cadence in most cases. Log slowly while the payload sits mid band, and switch to fast logging the moment the slope steepens or the door opens. You keep the shape of the events that matter and you keep the battery.