openJIIDocs
Devices & Protocols

Multi-device measurements

Run a shared measurement or route connected devices to family-specific cells.

A normal protocol cell runs the same protocol on every connected device in parallel. A workbook can instead use a connected-device branch to route each device to a different protocol or command cell. Each device still produces an independent result and upload; a shared workbook_run_id identifies readings from the same round.

The recording uses the web workbook host with mock devices. The mobile field flow uses the same per-device outcome model.

Set up a mobile round

  1. Connect several MultispeQs by USB. A powered USB hub is recommended.
  2. Open Devices and confirm the connected-device count.
  3. Start an experiment whose workbook contains the intended protocol.
  4. Start the measurement. The app runs the protocol on all connected devices.
  5. Review the results per device before accepting them.

A Bluetooth connection remains single-device, and the app does not mix a Bluetooth device into a wired multi-device setup.

Route by connected device

In a branch cell, choose the pinned Connected device source. Conditions can use family, id, name, zero-based index, or firmwareVersion. Point each path, including any default, to a protocol or command cell. At run time each connected device evaluates the branch with its own identity and runs only its matched target.

Devices that do not match a path, or whose path does not resolve to a measurement target, sit out that round while matched devices continue. If no device has a valid target, the workbook continues to the next cell. In the web workbook host, the identification handshake overrides a conflicting toolbar family and shows a mismatch notice.

Handle partial failure

If some devices fail, successful readings are retained. You can retry only the failed devices or continue with the successful results. Already-successful devices are not scanned again during that retry.

After acceptance, each reading is saved and queued separately. Offline retries therefore behave like ordinary single-device uploads; the shared run ID is correlation, not a batched upload.

Current limits: battery display tracks the first connected device, and workbook runs performed on the web are test runs rather than uploaded experiment measurements.

On this page