Skip to content

Follow an observation

Once you create an observation, the detail page is where you watch it run, see what came back, and get to the data. It’s organized as a header plus four sections in a sidebar on the left: Overview, Results, Data and Processing. This page covers the header and the Overview; the other sections are summarized below and have their own pages where noted.

Routes are /u/:slug/observations/:uid (user-owned) and /orgs/:slug/observations/:uid (org-owned), with /results, /data and /processing for the other sections. The short link /observations/:uid (optionally followed by /results, /data or /processing) redirects to the owner’s URL, which is handy for sharing.


The header band carries:

  • Name and status (active / completed / canceled).
  • Owner, and the creator when the creator is not the owner (the entity funding it vs. the user who authored it).
  • Share — the visibility (restricted or public) and the people and groups the observation is shared with. Any signed-in viewer can open it; only someone who can manage the observation can change it.
  • Edit — returns to the editor; see 5. Review and publish for what stays editable after you create it.
  • Notify (the bell) — turns notifications about this observation on or off for you.
  • More actions (the ellipsis): Re-process data (for people who can manage the observation), Resubmit as new observation (opens a fresh editor seeded from this observation’s spec), Download data… (see Download data), Export spec… (a portable JSON blueprint with identity, ownership and funding stripped) and Open in Afterglow… (see Afterglow).
  • Cancel on an active observation, or Activate on a canceled one, for people who can manage it.

Above every section, banners explain when the observation can’t run: when no instrument is eligible for it, or when its observing accounts are over quota (red when every account is over quota and nothing will be queued, amber when some still fund it). A status strip below them shows the observation’s progress and window.


The Overview reads top to bottom:

  • Where it stands — a one-sentence headline, a Needs attention box when something does (for example the window is about to close or the scheduler can’t place the work), and tiles for Delivered, Visits, Image quality, Telescope time and Next.
  • Latest image beside a Visits grid (one cell per visit, linking to that visit on Results), and Image quality over time.
  • Schedule & observability — Up next (the next visits: running first, then scheduled, then not yet placed, each with its telescope and time) and the elevation chart for the target. Once the observation has finished, this folds behind one line.
  • What you asked for — the data requests with their progress, the target, and a folded Configuration card (instruments, constraints, cadence and processing as submitted). This is read-only; to change anything material, head back to the editor.
  • Observing accounts — the accounts funding the observation, with their quota usage.
  • Activity — chronological events for the observation (the latest five, with “show more”), and the Projects the observation belongs to.

Visitors who are not signed in (on a public observation) see a Latest looks panel of recent images in place of the top part.


  • Results — one entry per sample: one image, or for a series request one run of images on one camera. Samples are planned when the observation is submitted and fill in as the telescopes capture them and the pipeline builds each result. Each entry shows one status (scheduled, capturing, processing, ready, interrupted, failed or not taken) and one quality line; chips narrow the list by request and by epoch, and a Gallery / Details toggle switches the layout. Clicking an entry opens that result’s own page, with the image viewer, the frames it was built from and its processing runs. Download results packages every result’s deliverable (see Download data). The individual tasks the scheduler created now appear under their results; see Tasks.
  • Data — every file attached to the observation, as large or small thumbnails or a details table, filterable by role, node QA, Skynet QA, storage location, file name and exposure window. Each file has its own page. The Downloads strip and the Download button live here. See Files.
  • Processing — the observation’s processing settings, with Save and a re-process button for people who can manage the observation. See Processing and analysis.

While an observation is active, or while a section still has work moving (frames transferring, a pipeline run queued or running), the page reloads its data every 20 seconds, pausing while the browser tab is hidden. A completed or canceled observation with nothing in flight stops refreshing.


A few things you might expect that live elsewhere:

  • Grant / account history — the Overview shows each funding account’s quota usage, but credit-usage history lives on the observing account and grant pages.
  • Telescope-side context — what else was running on the assigned telescope, what the weather was doing — lives on the telescope’s own page.
  • Project rollups — if the observation is part of a project, the project page is where you see cross-observation summaries.

  • Observation — the record powering this page.
  • ObservationResult — one sample of one request in one epoch; what the Results section lists.
  • Event — what drives the activity feed.
  • ObservationAsset — the file rows the Data section lists.