← Back to The Signal

Platform

How to configure a research wearable

By Sameer Sontakey

Configure sample rate, data cadence, active sensors, and raw export once across every device. Built for research and enterprise programs.

Quick answer: set sample rate, cadence, sensors, and raw export in a device profile before enrollment, then apply it once across every device.

Why configuration belongs before enrollment

A beautiful dashboard does not rescue a cohort collected at the wrong cadence. If your protocol needed high-frequency PPG and the devices wrote 15-minute summaries, you will not reconstruct the waveform later.

Consumer defaults optimize battery and a fixed app experience. Research and enterprise programs need the opposite posture: an explicit data model, applied consistently, with raw export available when analysis demands it.

The device profile

Sensor Bio Device Config is the Build layer. Think in four controls.

Sample rate

Example research-oriented setting: 100 Hz optical capture when your pipeline needs waveform detail. Lower rates may fit lighter program monitoring. Pick from the protocol, not from marketing defaults.

Data cadence

Cadence is how often data leaves the device path into your workflow:

  • Live
  • 1 minute
  • 5 minutes
  • 15 minutes

Live supports coaching and product features. Longer cadences reduce chatter for overnight or population programs. State the choice in the protocol.

Sensor toggles

Turn on only what the program needs:

  • PPG
  • Motion
  • Skin temperature

Unused channels still cost power and review time. Toggle with intent.

Raw export on or off

Raw export is the difference between rebuilding features later and being stuck with someone else's scores. Leave it on when models, validation, or re-analysis are in scope.

Together these controls are part of 100+ configurations you can set once for the organization.

Example profiles (program illustrations)

These are workflow illustrations for general wellness and research operations. They are not diagnostic protocols.

Overnight recovery monitoring

  • PPG on, motion on
  • Lower daytime urgency; prioritize overnight windows
  • Batch cadence acceptable if live coaching is out of scope
  • Raw export on if you will re-derive features

Algorithm and model development

  • PPG on at a high sample rate (for example 100 Hz)
  • Raw export on
  • Live or short cadence while you validate ingestion
  • Motion on for artifact context

Motion-forward activity program

  • Motion on, PPG on if heart-related program metrics matter
  • Cadence matched to how often coaches review
  • Raw export optional if you only need derived activity features

Rollout

  1. Draft the profile against the protocol, not against a sales slide.
  2. Apply the profile across devices for the org.
  3. Issue SDK and API keys from the admin console.
  4. Confirm streams in your app, warehouse, or notebook.
  5. Only then attach Signal Studio rules and adherence watchlists.

From config to action

Configuration decides what exists. Signal Studio decides who gets notified. Adherence watchlists show who needs follow-up on wear time, sleep, and activity without another survey.

For the API shape behind the streams, start at raw PPG access.

Claims note

Sensor Bio V2 is a general wellness device, not FDA cleared for diagnosis. Configuration examples are for research and program design. They are not clinical protocols and do not imply RPM billing eligibility.

FAQ

What can you configure on a Sensor Bio device profile?
Sample rate, data cadence (live through multi-minute batch), which sensors are active (PPG, motion, skin temperature), and whether raw export is on. Profiles apply across devices for your organization.
Why does sample rate matter for research wearables?
Sample rate bounds what algorithms and analyses can recover later. A profile that is too coarse cannot be fixed after enrollment if the raw signal was never stored.
Is this the same as a consumer wearable setting screen?
No. Consumer apps optimize for a fixed wellness experience. Research and enterprise programs need a data model you set once, then enforce across every wearer device.
How does configuration connect to alerts?
Configuration decides what is collected and how often. Signal Studio uses that signal to notify coaches or operators. Configure first, then add rules.