Skip to main content

Estimators

Estimators turn raw sensor signals into derived measurements that are more useful for experiments. They are similar to calibrations in lifecycle (created through a protocol, saved to disk, optionally set active), but they solve a different problem: they combine or transform measurements from multiple sensors into a single estimate.

When to use an estimator​

Estimators are useful when you want:

  • A derived signal that blends multiple sensors into one measurement.
  • A method that can be re-run when optics, hardware, or media change, without replacing core calibrations.
  • A stored, versioned artifact that the Pioreactor can reuse across experiments.

Example: fused optical density​

The od_fused estimator combines the 45°, 90°, and 135° optical density sensor readings into a single fused OD value. This estimator is created by the OD fusion protocols and saved to disk so future experiments can use the fused signal automatically.

Where estimators live​

Estimators are stored on each Pioreactor at:

$DOT_PIOREACTOR/storage/estimators/<device>/<estimator_name>.yaml

The active estimator for a device is tracked in local persistent storage. Estimators are not listed in the Calibrations UI or CLI, but they are visible in the Estimators page in the UI.

Managing estimators in the UI​

Estimators page showing filters, table, and active toggle.

Use the Estimators page to:

  • Filter by Pioreactor or device.
  • See which estimator is active for a device.
  • Focus on active estimators only using the toggle.