Butterfly Effect

University / academic research group

A collaboration model for longitudinal sleep research, transport, and interpretable physiology.

Best suited for academic groups with longitudinal cohorts, repeated outcomes, and interest in transport, harmonization, and publishable validation studies.

Why this institution is a strong fit

Universities are a strong fit when they can contribute well-annotated longitudinal cohorts and a clear research question. The platform already supports open harmonization, route governance, and portable-family probes, so the collaboration can move quickly toward a publishable methodological or applied study.

Longitudinal cohortsPublishable methodsCross-cohort validation

What is already defensible

Night stress
R² 0.238
Night D+1 stress
R² 0.346
Event stress portable
R² 0.316
Main gap
Night transport

Institutional fit

Academic datasets can turn the current system into a stronger research program

The strongest academic contribution is a cohort that adds methodological leverage: comparable outcomes, longer follow-up, or richer physiology aligned with repeated self-report.

What we bring

  • A reproducible data-to-report research pipeline
  • Two nightly routes that now clear promotion gates under explicit governance
  • Open-cohort baselines for comparison
  • Portable-family and route-level validation logic
  • A clear separation between accepted, experimental, and support-only claims

What you contribute

  • Longitudinal cohort with repeated outcomes
  • Consistent annotation and study protocol metadata
  • Wearable or sleep-derived physiology
  • A concrete research question suitable for a pilot manuscript or grant work package

Expected outcome

  • Cross-cohort transport evidence
  • A stronger portable-family study for stress or non-stress targets
  • A methods paper around interpretation, harmonization, or generalization
  • Groundwork for grants or follow-on prospective studies

Data package

What an academic cohort should contain

Longitudinality and outcome quality matter more than raw modality count alone.

Minimum required

  • Nightly sleep features over weeks or months
  • HR/HRV or other nocturnal physiology
  • Repeated PROs/EMA for symptoms or burden
  • Study metadata, inclusion criteria, and confounders
  • Stable identifiers and dates for longitudinal linkage

High-value expansion

  • Raw PPG / IBI / accelerometry
  • Additional clinical instruments with clean timing
  • Daily self-report or event-linked outcomes
  • Small PSG-anchored validation subset
Key point: For academic collaboration, the best datasets are not the noisiest or largest by default; they are the ones that let us test the right methodological claim clearly.

Pilot model

Recommended pilot for a university partner

The ideal academic pilot is designed to answer one publishable question cleanly.

Recommended first pilot

  • One cohort, one target family, one transport or interpretation question
  • Retrospective harmonization against the open stack first
  • Transparent held-out and subgroup analysis
  • Joint authorship and protocol framing from the beginning

Outputs to the institution

  • Technical harmonization package
  • Portable-family or route-level benchmarking
  • Draft figures/tables for manuscript or grant use
  • A recommendation on whether to expand prospectively