Automatic runs

Instead of asking for each run, give the satellite a schedule once and keep uploading its fixes as they come down:

client.schedules.set(satellite)  # the defaults
client.schedules.set(satellite, arc_hours=30, setting="segments-60min")  # change some
client.schedules.get(satellite)  # its settings; None without one

With the default settings the service determines one orbit a day: each arc is a UTC day with an hour before and after it, so that neighbouring arcs overlap by two hours.

  • An arc is run as soon as your data reaches its end. If the data stops short of the arc’s end, the service waits for more: max_wait_hours (12) after the arc’s end and after your last upload it runs the arc with what there is. An arc without any data is passed over once later data exists.

  • Data that arrives late is not lost. When you upload fixes for a time the schedule has already passed, it goes back to the first arc that holds them and runs the arcs from there again, in order. This reaches 30 arcs back; for older data, set first_arc_start.

  • Each run continues the one before it: it starts from that run’s orbit. Its estimate still rests on its own arc’s data alone.

  • The latest accepted run is the satellite’s orbit. A run that is not accepted is kept, and the orbit before it stands.

Setting

Default

Meaning

enabled

True

False stops the automatic runs; the settings and the place are kept

arc_hours

26

Length of each arc

interval_hours

24

From one arc’s start to the next one’s

max_wait_hours

12

How long an arc with incomplete data waits for more

prior_covariance

False

True: each run also takes the previous run’s uncertainty at its start as prior knowledge, which ties the arcs together

first_arc_start

the arc that holds today

Where to begin. Give a past time to work through data you have already uploaded: upload it first, oldest first

setting, custom, profile

the profile’s default

As for a run (Orbit determination runs). Custom settings stay until you send custom=None or name a setting

Setting the schedule again changes what you give and keeps the rest. Runs started by the schedule have trigger == "schedule", and do not count against your limits.

The satellite’s orbit

orbit = client.orbits.get(satellite)  # None before the first accepted run
print(orbit.od_run_id, orbit.ephemeris_start, orbit.ephemeris_end)
client.orbits.ephemeris(satellite, "orbits/")  # its CCSDS OEM: the arc and the day after

The ephemeris changes only when a new run is accepted (orbit.ephemeris_sha256 with it), so asking for it often is cheap. To be told instead of asking, use a webhook (Webhooks).