Eight FUAs · 76 annual H06 observations each · 608/608 QA.
From a working evidence engine to territorial intelligence.
The V0 changed the roadmap. The question is no longer whether we can build a reproducible territorial trajectory. We can. The next question is how far that evidence can be extended, interpreted and validated before we move from observation to outlook.
PM2.5 three-year panel; PM10 2022–2023 with historical backfill underway.
Data-backed Territory Intelligence views running from the same validated bundle.
Rebased roadmap
V0 · Evidence → Territory
Stabilise the eight-FUA product: H06 1950–2025, PM2.5 2022–2024, PM10 2022–2023, provenance, QA and data-backed Territory Profiles.
V0.2 · Evidence depth
Backfill PM10 and PM2.5 history; add a small number of high-value physical and territorial metrics; resolve commercially usable geography.
V0.3 · H-D-A-R evidence map
Operationalise Habitability, Desirability, Affordability and Resilience as distinct user lenses while keeping scientific metric roles separate and explicit.
V0.5 · Evidence-first scale
Expand each admitted phenomenon to its maximum defensible source coverage; test geography sensitivity, missingness, comparability and Product Eligibility.
V0.7 · Evidence governance
Operationalise Depth × Quality × Coverage, Claim Governance and fail-closed comparison before broad autonomous scaling.
V0.9 · Scenario prototype
Introduce explicit 10/20/30-year scenarios with uncertainty only after historical evidence and backtests are credible.
Habitability Intelligence
A governed territorial evidence system joining H-D-A-R, State / Trajectory / Evidence, Fingerprint, Compare and grounded Ask at the broadest defensible scale.
Evidence → Governance → Trajectory → Interpretation → Compare → Ask.
We do not need a universal score to create value. The primary product object remains the Territory Intelligence Profile; composite scoring is optional and remains gated by empirical validation.
What we should optimise for now
Ingest broadly, compute systematically, publish selectively. Scale evidence by source coverage while keeping provenance, rights, geography, uncertainty and claim boundaries explicit.