Skip to content
PilotWorking Public Beta

How our signals are built

This page explains how Phuket Local Connect collects, labels and reviews the Phuket Well-being Signals — and the limits of what a founder-led public beta can honestly claim today.

Signal types

  • Community observations — Recorded by residents or participants — coastal, wildlife or waste sightings. Not automatically confirmed science.
  • Public reports — Via Report & Connect. PLC coordinates where possible; it is not an emergency service.
  • Third-party open data — Such as iNaturalist, DMCR, GISTDA. Source and last-updated date are always shown.
  • Participant reflections — Published only in aggregated / anonymous form to protect participants, especially minors.

Five evidence types

A type says where information comes from; it is not a quality score. These categories must not be visually conflated.

Official source
A direct or linked record from the source owner. Showing it is not an endorsement by or of PLC.
PLC calculation
A transparent calculation PLC made from a cited source. The method is shown.
PLC field evidence
Observations or data collected through a PLC method, with scope, date and method shown.
Community signal
Submitted by people and not verified. A signal, not a fact.
PLC interpretation
A question or synthesis by PLC. Not a source fact.

Observation is not automatically fact. Activity is not automatically outcome.

Outcome claim states

Planned

An intention, not evidence that it happened.

Observed

A recorded observation within a stated scope.

Measured

A stated method, baseline or comparator, and period support the measure.

Independent verification documented

An identifiable independent reviewer checked the evidence using a disclosed method. This does not replace PLC’s eight-stage evidence label.

PLC does not use ‘verified impact’ unless an independent review actually occurred and the reviewer, method, scope and limitations are shown.

The 8-stage evidence standard

Public records use precise evidence stages instead of one broad "verified" label. Each stage says what actually supports a record; it is not a quality score.

  1. 1Submitted — Someone submitted it; not yet checked.
  2. 2Screened — PLC reviewed it before listing; not a confirmation of fact or by an agency.
  3. 3Source-backed — Linked to a named source.
  4. 4Community-confirmed — Confirmed by more than one local contributor; not an official fact.
  5. 5Specialist-reviewed — Reviewed by an identified specialist.
  6. 6Agency-confirmed — The responsible agency confirmed in writing.
  7. 7Action in progress — A recorded action is underway; not yet an outcome.
  8. 8Outcome measured — An outcome was measured with a stated method, not just activity.

Reviews are performed by the founder and authorised reviewers — this is not formal scientific certification.

Source provenance

Every data item should show its source owner, link, dataset date, PLC check date, geography and limitations. Government remains the source of record, and no figures are shown until a validated retrieval succeeds.

Outputs vs outcomes

Activity is not outcome. Attendance, items collected or forums held are outputs. An outcome needs a baseline, a method and evidence before it can be shown as "Outcome measured".

Community, agency and specialist distinctions

  • Community-confirmed: several independent local confirmations; not an official fact.
  • Specialist-reviewed: reviewed by an identifiable specialist using a stated method.
  • Agency-confirmed: requires a direct official source; forum discussion is not a government commitment.

Uncertainty and missing data

We label uncertainty and missing data rather than inventing precision. Stale, community-reported or modelled information is never shown as current official fact.

Provenance, public vs restricted, and corrections

Provenance recorded: source owner, link or reference, source date, observed date, PLC review date, method, reviewer type, confidence (high / medium / low / unclear — a label, not a score) and limitations.

Only reviewed items chosen for publication are public. Client project evidence and unreviewed items are private by default. Sensitive locations are shown at area level only.

A correction creates a new item that supersedes the old one; history is kept. To request a correction or challenge an item, use the Rights & Corrections page.

AI may help translate or summarise, but is never the final verifier. People decide evidence stage and publication.

Evidence readiness is an internal completeness view — not a certification, assurance, compliance finding or GSTC/TNFD score.

Rights & Corrections

How PLC fits with existing systems

PLC complements official city-data and specialist systems. It connects source-aware local evidence, participation and follow-up; it does not replace government data platforms or operational agency systems.

PLC is a place evidence and participation layer — not another biodiversity database.

PLC complements public-agency databases, weather and water services, geospatial systems, city-data platforms, research databases and emergency-warning channels. It does not replace their ownership, mandate or official channels. Referencing a source does not imply partnership or endorsement.

Authorised agency staff can send an update or correction with an official source. Submissions are private and never change an evidence stage automatically. Agency update

Source Registry

Sources PLC references, with owner, access mode, check status and limitations. The owner remains the source of record.

A listed source is not a partner or endorser of PLC.

Loading…

    Metadata and interoperability

    • Metadata first: owner, dates, method and limitations before any figure.
    • Stable IDs, timestamps and source lineage; material corrections keep history.
    • Machine-readable export comes before any public API.
    • Designed to be compatible in future with DCAT-like dataset metadata, STAC for geospatial/time assets and GeoJSON/OGC exchange — not yet tested, and no compliance is claimed.

    Limitations of this pilot

    • PLC is not a government agency or emergency service.
    • Not a Nature Positive claim or ESG certification.
    • Participation counts are not the same as ecological outcomes.
    • All numbers carry source, updated date and confidence level.

    Example framework — not certification

    PLC Evidence Receipt — example framework

    This receipt structures a record of participation and findings. It is not certification, an audit, independent verification or an endorsement by PLC.

    Who participated

    Participant group and role — recorded with consent

    What was done

    Activity and method — linked to a dated record

    Place

    Named Phuket place or ecosystem where appropriate

    What was observed or measured

    Source, method, date and limitations shown

    Evidence status

    SubmittedSource-backedSpecialist-reviewedOutcome measured

    This example shows Planned only. Later statuses apply only when evidence supports that level.

    What cannot yet be claimed
    No ecological, social or sustainability outcome is verified by this planned record.
    What happens next
    Run the activity, record sources, review the evidence and update the status only when supported.

    Paying Regenera Thailand does not buy Phuket BiodiverCity coverage, PLC verification, ranking, endorsement, certification or guaranteed positive sustainability claims. Payment does not change an evidence status, and personal data is not sold.

    Rain / Flood