Currently have Blue Iris, Frigate and HA set up.

Im using the Frigate integration occupancy sensor to trigger a REST command to tell blue iris to record. After 5 seconds of no occupancy, another REST command is sent to stop recording.

snippet from my config.yaml is below:

####DRIVEWAY###########################
  driveway_zone_a_frigate_trigger: 
    url: http://192.168.0.100:81/admin?camera=driveway&trigger&zones=1&memo=frigate_triggered_motion

  driveway_zone_b_frigate_trigger: 
    url: http://192.168.0.100:81/admin?camera=driveway&trigger&zones=2&memo=frigate_triggered_motion

  bidriveway_frigate_stop_trigger: 
    url: http://192.168.0.100:81/admin?camera=driveway&trigger=0&memo=frigate_stop_triggered_motion

####DOORBELL###########################
  doorbell_zone_a_frigate_trigger: 
    url: http://192.168.0.100:81/admin?camera=doorbell&trigger&zones=1&memo=frigate_triggered_motion

  doorbell_zone_b_frigate_trigger: 
    url: http://192.168.0.100:81/admin?camera=doorbell&trigger&zones=2&memo=frigate_triggered_motion

  bidoorbell_frigate_stop_trigger: 
    url: http://192.168.0.100:81/admin?camera=doorbell&trigger=0&memo=frigate_stop_triggered_motion

This set up is working flawlessly, except but I'm occasionally seeing some triggers aren't stopped and the camera continues recording until the next frigate motion event ends.

My automation is basically

If frigate occupancy sensor for {camera} is none for 5 seconds, send rest command {stop trigger}

Is there a way to make that more fool proof? maybe a 2nd check for no occupancy 10 seconds later?