Follow-up to #577.
The current /alerts/export endpoint only exposes 5 columns (id, lat, lon, started_at, last_seen_at), which is too thin for studying past alerts.
We want the export to include every field relevant to alert analysis.
Proposed columns
Core
alert_id
started_at_date (UTC)
started_at_time (UTC)
last_seen_at (ISO)
duration_seconds
lat, lon (empty when null)
organization_id
Sequences (multi-value cells, |-separated, ordered by sequence.started_at ASC so positions align across columns)
sequence_ids
sequence_count
sequence_labels (is_wildfire: wildfire / other / unknown)
sequence_azimuths
pose_ids
Cameras (same ordering as sequences)
camera_ids
camera_names
camera_count
Follow-up to #577.
The current
/alerts/exportendpoint only exposes 5 columns (id,lat,lon,started_at,last_seen_at), which is too thin for studying past alerts.We want the export to include every field relevant to alert analysis.
Proposed columns
Core
alert_idstarted_at_date(UTC)started_at_time(UTC)last_seen_at(ISO)duration_secondslat,lon(empty when null)organization_idSequences (multi-value cells,
|-separated, ordered bysequence.started_atASC so positions align across columns)sequence_idssequence_countsequence_labels(is_wildfire:wildfire/other/unknown)sequence_azimuthspose_idsCameras (same ordering as sequences)
camera_idscamera_namescamera_count