01
Design studies
Record the research question, hypothesis, objectives, sampling plan, methods, inclusion rules and statistical plan.
S.W.A.N. Research
From acoustic detections to defensible research.
The protected research platform goes beyond browsing records. Researchers can define protocols, assign control and intervention stations, preregister methods, freeze versioned datasets, run exploratory analyses and preserve a complete audit trail.
Scientific guardrail
Automated detections are observations—not abundance estimates.
The platform keeps confidence, recorder coverage, station imbalance, classifier limitations, privacy controls and reproducibility information visible throughout the workflow.
02
Preregister protocols
Freeze a timestamped, fingerprinted protocol version before inspecting or interpreting the final results.
03
Build datasets
Create immutable snapshots with exact filters, variable definitions, quality reports and SHA-256 fingerprints.
04
Analyse transparently
Retain parameters, assumptions, quality warnings, outputs and a reproducibility scaffold for every analysis.
Analysis laboratory
Useful exploratory tools, with limitations shown beside the result
Descriptive network and station profiles
Species richness, Shannon, Simpson and evenness indices
Daily, hourly, monthly and weekday activity patterns
Control and intervention cohort comparisons
Before-and-after exploratory effect estimates
First and last detection phenology summaries
Dataset quality, imbalance and sample-size warnings
CSV exports and external-analysis reproducibility code
AI research copilot
Interpretation without handing AI the raw archive
When enabled, the copilot receives the study design, aggregate statistical results and quality warnings. It does not receive raw audio or the complete detection table.
It drafts cautious summaries, identifies unsafe claims, highlights confounding and suggests checks. Every response remains subject to human review and is retained with the analysis.
Controlled access
Detailed data is shared responsibly.
Applications are reviewed so that sensitive location context, private audio states and automated classifications are used with appropriate safeguards.
1. Apply
Explain your organisation, research purpose and intended data use.
2. Review
S.W.A.N. checks that the requested access level is suitable.
3. Research
Approved accounts receive the protected study and evidence workspace.