The planet scale is the forcing layer. Earth foundation models — Aurora, GraphCast, GenCast — provide the atmospheric and climate signals that PLANETAI reads downward through bioregion, city and community tiers. Six planetary boundaries set the outer frame; four pilot bioregions are where measurement becomes action. Toggle layers to explore what the instrument is already tracking.
Planet · production → place
What a country buys, how those products convert into materials, and what it throws away — the planetary scale of the production-place relationship. Imports from the Atlas of Economic Complexity (2023) on the left; municipal solid waste from the World Bank What a Waste 2.0 on the right. Between them sits the editorial piece no one publishes: a crosswalk that converts product categories into material fractions. We publish ours — it is a contribution of this observatory, not a measurement, and we say so.
Imports · AEC 2023 · USD billions by HS section
Editorial bridge — product→material mapping
Crosswalk derived by FCF (April 2026) from Eurostat EW-MFA (biomass / metal ore / non-metallic mineral / fossil), OECD Global Material Flows, and Krausmann et al. (2017). No single authoritative public HS2→MSW crosswalk exists. Ours is open for revision.
Waste · WaW 2.0 · kg/capita/year by MSW fraction
The planet view overlays the Fab City network onto live planetary feeds: Earth-system climate indicators (Open-Meteo / Aurora at city locations), real-time air and maritime movement, active fires, seismic events, and biodiversity signals. Each layer above is a discrete toggle — some pull live on click, some use bundled baselines. This is where PLANETAI becomes an instrument rather than a roadmap.
Global mean anomaly, ENSO, wind — runtime overlay on live host.
Regional pollution + aerosol fields at city centroid for the year.
Species occurrence records logged worldwide that year.
The bioregion is the ecological frame at which a metropolitan area meets its supply chain — the watershed, the foodshed, the materials shed. Each pilot anchors to a regional / state-level open-data publisher (Generalitat de Catalunya, Massachusetts state, Metropolitan Region of Santiago, Bali Province) and a fab-lab / Fab City consortium on the ground. Switch pilot; everything below re-reads.
Bioregion · production → place
What this bioregion consumes, what it converts into materials, what it lets go of. This is the Fab City core principle made visible at metropolitan scale. Barcelona and Boston anchor to published metropolitan and state-level waste accounting; Bali and Santiago don't yet — those Sankeys are synthesised from national per-capita profiles and flagged. Closing that gap is what the four pilots are for.
Bali · sovereign open data ingest
Bali Satu Data (provincial portal) federates upward into the national Satu Data Indonesia portal under Pergub Bali 53/2021. The cards below are the discovered datasets per tier, ingested under their native Bali Satu Data taxonomy. The Tri Hita Karana ↔ four-pillar mapping is a Bali pilot only co-design experiment (Path B) gated on CAST + MDG + IT Del + a named Balinese customary authority sign-off; it is not applied at ingest by default. Discovery only — values flow once the resolver wires through to each publisher's source URL.
AQI · 24h profile · 500 m · satellite + weather + traffic fused
Per-capita municipal waste, Barcelona, last 10 years · source: Generalitat 69zu-w48s
H₀-A observable: negative residual = community action visible at sub-500m scale.
Fab-lab catchment footprint change 2020→2023 · physical capacity proxy for material throughput upper bound
PLANETAI runs in four bioregions: Europe (Barcelona), North America (Boston), Latin America (Santiago de Chile), Asia-Pacific (Bali). Each combines a different climate, sensor density, partner consortium, and regulatory regime. SmartCitizen devices provide proxy air quality readings; colocation calibration against co-located FRM/FEM reference instruments (via OpenAQ) is required per bioregion site before Aurora residuals can be attributed to distributed production. The picker above the header switches the map and local live data panel.
Bioregion · river-basin flood risk · Google Flood Hub
A flood event is a material throughput shock — it disrupts the Environmental-pillar material flows at the scale where a Fab Lab network can actually respond. Google Flood Hub provides 7-day riverine warning for 100+ countries; all four PLANETAI pilot river basins are covered. Barcelona's Llobregat/Besòs was the closest active flood system to the 2024 Valencia DANA event — exactly the hazard class Flood Hub was designed for.
The headline metric for Fab City progress, benchmarked against the 2054 Barcelona Pledge (50% self-sufficiency, signed at FAB10 on 6 July 2014), and the falsifiable target for PLANETAI's hypotheses. One composite score, four pillars (Environmental · Social · Economic · Governance), every territory.
Built from the Full-Stack Metrics 4-Pillar × 5-Scale framework, the v0 strap weights pillars per Boeing 2024 (NACE × COICOP sectoral self-sufficiency + circular-economy distance) — Environmental 0.30 · Economic 0.30 · Social 0.20 · Governance 0.20. The next wave layers PITO/DIDO and the response coefficient ρ on top so the score reads as FCIt = DIDOt · (1 − PITOt) · ρt, with the trajectory ΔFCI/Δt measured against the pledge curve. Pick the city, pick the axes, overlay a comparison. The number on the right is what a mayor, a community council, or a Fab City signatory acts on; the strip beneath it shows distance-to-pledge. The radar / bar chart below decomposes how the score was reached.
—
Pick two axes. The eight reference cities place themselves in the resulting 2D space. This is the analytical surface — not just the snapshot, but the relationships between what the four pillars carry.
A single high correlation between two axes is descriptive, not causal. Use this view to surface candidate relationships, then verify against the source connectors. The four pilots (Barcelona · Boston · Santiago · Bali) carry brand colours; the four reference cities (Amsterdam · Paris · Shenzhen · Detroit) carry a neutral grey for visual contrast.
The same cross-plot pattern, but now per municipality within the AMB and from real open-data sources. Picks 14 AMB municipalities + Barcelona; joins on municipality name + most recent year; computes Pearson r on the live values.
This is the analytical move PLANETAI claims: federate real open data across municipalities, place every unit in a comparable space, and let the relationships themselves be the question. The instrument is what funds the wire — the same pattern lands for Boston / Santiago / Bali as MCP-equivalent layers ship.
Satellites measure in tens of kilometres. Neighbourhoods live in metres. The locality PLANETAI measures is the place where you know the people around you — the few square kilometres of street, market, school and workshop you actually move through. The fab lab sits inside it as one tool among many; the unit of measurement is the neighbourhood, not the lab. Weng et al. (2024, npj Urban Sustainability) frame this gap as the turn from machine-pixel vision to “people as virtual sensors” — and admit that public training datasets focus mainly on developed countries. Fab City's distributed network answers from the opposite direction: community-sensed hardware shipped from Barcelona since 2013, with node density weighted toward Yucatán, Thimphu, Shenzhen and Curitiba.
api.smartcitizen.me/v0/devices/world_map.
LIVE
PLANETAI measures at five scales (Vivanco's Full Stack Metrics Framework). Give permission to use your location and the instrument shows you where you stand in each — the planet-tier data above you, the bioregional ecology around you, the regional / state-level governance and supply-chain layer, the city you live in, and the community-scale fab-lab and sensor network at your neighbourhood.
Your coordinates stay in your browser. Used only to query PLANETAI data sources at your location; not stored, not transmitted to PLANETAI servers.
The hyperlocal scale of the production-place relationship — the 2–5 km radius of your block and the people around you. Nobody measures the material budget of a single neighbourhood at this resolution yet, and that is precisely what Weng et al. flag as the gap planetary observation cannot close. This Sankey publishes the shape of a neighbourhood material ledger: a 410 kg/cap annual household throughput split across six consumption categories and routed to six disposal fates. The wiring exists; the weighing doesn't. The next phase weighs this — at the neighbourhood, not at the fab lab.
Neighbourhood data inventory · what we measure when we measure here
The neighbourhood is invisible in formal statistics. Trash collection happens but isn't published per block. Air quality is monitored at a few city stations, not at your corner. Repair shops and informal vendors don't appear in any registry. PLANETAI's job at this scale is to fill the inventory — through community contribution, dedicated data harvesters (image / scrape / signal pipelines), and extensions of the existing community-sensor stack. The list below is what the 36-month measurement window builds.
Volume and composition of what gets put out, when. The smallest unit at which the production-place loop becomes visible — and the data nobody publishes.
PM2.5, NO₂, CO at 50 m – 1 km grain. The corner you breathe at, not the ministry station 8 km away.
Daytime + night-time dB by hour and place. The Plaça-del-Sol pattern: the campaign that produced policy.
Surface-temperature variation between blocks: shade, paving, tree cover. 30 m Landsat + cheap ground-truth thermometers.
Where the street floods after rain, where pipes leak. Photo-based citizen reporting, geotagged, classified.
Per-block tree count, shade fraction, species diversity. The thermal regulator nobody publishes at this grain.
Per-building suitability for PV: orientation, shading, area. The pre-condition for a distributed energy transition at the block.
Birds, insects, urban wildlife at street resolution. Where Sparrows still nest, where pollinators forage, where nothing lives.
Shops, clinics, schools, pharmacies, transit within 5–15 min walk. The neighbourhood as a complete-life unit.
Workshops, libraries of things, fix-it cafes, Precious Plastic nodes. The infrastructure of distributed repair.
Plazas, parks, civic centres people actually use. Where the neighbourhood happens, not just where it's zoned to.
What languages are spoken on signs, in shops, on walls. Cultural granularity formal census misses.
Curb cuts, ramps, crossings, broken pavement. Whether a wheelchair, a stroller, or a tired body can move through here.
Distance from each block to the nearest fresh-produce vendor. The signal that distinguishes a fed neighbourhood from a fed-on-packaging one.
When plazas fill, what for, who shows up. The activity pattern that distinguishes designed-for from used-by.
What's made, repaired, recycled inside the neighbourhood. Fab-lab + workshop logs, weighed at intake and output.
Consumption per household and small business at the smallest aggregation utilities will publish.
Where the neighbourhood's groceries actually come from. Mercabarna-class wholesale + corner-shop receipt sampling.
Street vendors, micro-shops, gig work hubs. Invisible to tax registries; ~75% of urban employment in Santiago + Bali.
Pavement quality, lighting outages, blocked drains. The maintenance load no formal map captures by block.
Urban gardens, community-supported agriculture, balcony harvests. The shed-scale layer of the food system.
Litres per day at the smallest aggregation utilities will publish. The lever for distributed water re-use planning.
How many residents voted, what they prioritised, what got built. The closest thing to neighbourhood-scale democracy.
Meetings held, decisions logged, attendance. The functional health of the neighbourhood-scale veto PLANETAI architects around.
What residents are reporting (potholes, lights, noise, dumping). The neighbourhood's pulse, in maintenance requests.
This is what "the last mile" means. The gap between the planetary forecast above you and your own breath is what PLANETAI exists to bridge. Each tier has its own data, its own governance, its own sovereignty boundary. If you want the scientific spine — the scale cascade, the coverage matrix, the bridge table, the four research questions — it lives in the Understand tab. This page keeps its focus on where you sit inside the instrument.
Why local matters: inundation risk at the neighbourhood scale shapes the lives of the people around you — material logistics, sensor deployment windows, and the continuity of every workshop and lab in the affected blocks. Flood Hub's 7-day probabilistic forecast at 90m resolution is the earliest publicly available signal at this grain.
A 0–100 score measuring how close a city is to producing what it consumes — and how quickly its community can translate what it observes into what it does.
FCIt = DIDOt · (1 − PITOt) · ρt — Full-Stack Metrics 4-Pillar × 5-Scale [Vivanco 2024 / 2025]
The Full-Stack Metrics 4-Pillar × 5-Scale framework places every indicator the instrument collects into one of twenty cells — four pillars (Environmental, Social, Economic, Governance) at five scales (Planet, Bioregion, Region, City, Community). Each cell publishes its primary indicator family, its source, and a live / mock / research status. No floating indicators. If a cell has no source, it is named as a deferred research deliverable rather than glossed.
Cells outlined in blue are the Fab City Index lineage. At City × Economic, PLANETAI is the third generation of the Index — extending Florentin/Chabanel/Guimas (Utopies & FabCity Paris 2018) and Boeing (Hamburg 2024) into the federated four-pillar matrix.
Most phenomena downscale better than they upscale. Community observations rarely reconstruct planetary reality; planetary models rarely resolve community reality. That asymmetry is the instrument's central problem.
Show full bridge table →The scientific contribution of PLANETAI is the bridge, not the sources. For each phenomenon, two questions: (downscale) how much of community-scale reality can we reconstruct from planetary-scale inputs? (upscale) how much of planetary-scale reality can we reconstruct by aggregating community-scale observations? R² values are indicative targets — Experiment E measures the first one for PM2.5.
Query: "Find global locations with material-use and land-cover profiles similar to the Fab Lab Barcelona catchment (Poblenou, 41.40°N 2.20°E, 500 m radius)." AlphaEarth Foundations runs Sentinel-2 embeddings at 10 m resolution. Matched sites are places where PLANETAI's production-visibility gap is globally recurrent — not a Mediterranean anomaly.
The red cells in the coverage matrix above are the gaps. The Feed tab lists every connector — live, mock, and planned — that fills them. Each connector maps to one or more cells in the 4×5 matrix.
Feed · what the instrument ingests
An observatory is only as honest as its inputs. Feed lists all thirty-one connectors by layer — environment, production, knowledge — plus the paths for extending the set. Live connectors carry endpoint + auth; the rest carry SYNTHETIC · CONNECTOR PENDING until their data arrives. Tokens live only in your browser session.
Layers 2 and 6 — the environmental baseline the instrument reads against and the bioregional hazards / biodiversity that define where it reads. Each card shows its endpoint, its auth model, and what it contributes. Tokens live in session storage only — cleared when the tab closes.
Layers 1 and 3 — what can be made, where, with what equipment, and how matter flows through the system. These connectors ground the distributed-production estimand.
Layers 4 and 5 — the design commons the instrument indexes, and the spatial / logistics / movement feeds that describe how people, goods, and signals traverse the bioregion.
PLANETAI is open-source and the instrument only gets better when the network sharpens it. Four paths — report something broken, propose something missing, fork and adapt it to your city, or read what is real today versus placeholder.
Something rendering wrong, a broken endpoint, a claim that does not match the data. Short reproduction steps help.
Open an issue →A data layer we are missing, an axis worth adding to the Constellation, a scale bridge the instrument should measure.
Propose it →The full source lives on GitHub. Clone it, adapt it to your territory, open a pull request when it is ready to feed back.
github.com/fab-city/planetai-observatory →What is real right now, what is pre-fetched, what is still placeholder. Honest declaration — iteration 6, April 20 2026.
This platform is open-source under Apache 2.0. Fork it, adapt it to your city, submit improvements back.
Act · the operational chain
Observation alone doesn't change anything. These five verbs are a chain: Connect (find your network), Measure (instrument your place), Learn (read what others built), Design (adapt it for your context), Make (fabricate and deploy). Each step runs on a different tier of the Fab City network. The response to a measurement becomes the input to the next one.
Earth-observation tools (Copernicus, Earth Engine, smart-city dashboards) feed policy cycles that move on years-to-decades timescales. PLANETAI targets the bioregional tier — the scale at which a Fab Lab network can actually fabricate a response before the problem has moved on.
Written as design targets, not outcome claims — each arc shows what the observation-to-action chain looks like when every tier works. Specific durations are placeholders until the C2, C4, and C5 connectors land.
Open-source is not decorative here. A bioregional response chain that depends on proprietary hardware, closed firmware, or licence-gated data cannot be fabricated locally, cannot be audited by a community council, and cannot be re-run when the problem recurs in another bioregion.
PLANETAI's taxonomy is built around open-source hardware (OSHWA), open-design manifests (OKH), open spatial infrastructure (OKW, OpenStreetMap), and open scientific sources (OpenAQ, Aurora, Earth Engine open collections). The action layer is only as fast as the slowest closed licence in the chain — so the chain is built open by construction.
The observatory is a network before it is an instrument. Pledged Fab Cities carry a consortium; fab labs carry a workshop; other maker initiatives carry parallel communities. Your entry point.
Your coordinates stay in your browser. Used only to find the nearest pledged Fab City and Fab Lab; not stored, not transmitted to PLANETAI servers.
Grant location permission above to load nearby Fab Labs.
Measurement in PLANETAI is not satellite work. It is the neighbourhood — the block, the barrio, the 2–5 km radius where you know the people around you — instrumented by the people who live inside it. The fab lab sits inside that neighbourhood as one tool among many. The Local tab shows what that instrumentation looks like today: a live Smart Citizen layer of community-sensed environmental data, anchored on the same open-source hardware stack IAAC has shipped since 2013. The citizen-observatory toolkits below, developed with Fab City Foundation, are the operational handbook for turning a community into a measurement institution. Deploy a kit, join the Fab Academy Data Champions pathway, and you become a node in the PLANETAI network.
Methods and protocols for designing a monitoring programme with your community. Covers sensor selection, calibration, siting, and data-quality protocols.
Open toolkit →How to make a citizen observatory actually inclusive — not just inviting. Recruitment, retention, and equitable participation protocols for communities that are often spoken for rather than spoken with.
Open toolkit →The setup phase. Co-designing the observatory with the community you'll be measuring alongside. Scoping, objectives, partnership formation, and governance.
Open toolkit →Turning measurement into action. How observatory findings become policy proposals, infrastructure decisions, and public-facing evidence — the part most citizen-science projects never reach.
Open toolkit →Toolkits from the CITIOBS EU H2020 project (citiobs.eu), with Fab City Foundation as a contributing partner.
The Fab ecosystem is also a learning ecosystem. Fab Academy, Fab City Summit, and a handful of adjacent open curricula are where you pick up the craft before you pick up the soldering iron.
Link directory curated April 2026 — submit a PR on the platform repo to add yours.
Thousands of open hardware projects are documented, certified, or searchable through global open-design registries. PLANETAI surfaces these as a live layer — so a maker in your bioregion can find projects designed to be remixed, forked, and locally manufactured. Two registries below: OSHWA certified products, and the Open Know-How (OKH) open-design index.
Search a static snapshot of the OSHWA certification registry. Text searches project name and description; country and type dropdowns filter client-side.
Static snapshot of 50 OSHWA-certified projects, April 2026. Live API integration in v1.1 (requires free API key from certificationapi.oshwa.org).
The OKH standard — maintained by the Internet of Production Alliance — indexes
open-hardware projects that publish an okh.yml manifest. Search a
curated snapshot of ~30 well-documented OKH-indexed projects; results render
below. External link per card.
Snapshot of ~30 well-documented OKH-indexed projects. Live search at github.com/iop-alliance/okh-search (external).
Open collaborative hardware platform. Filter searches by location to discover projects published by makers in your bioregion.
Open Wikifactory project search ↗Some external links may 404 if a registry has moved a page — all registries above are also reachable from their root URLs.
Designs become things at workbenches. The fab labs you found under Connect are the first call; below are two further paths — verified regional manufacturers (where data exists) and local repair/crafts networks sourced from OpenStreetMap.
v1.1 integrates Make.Works API for verified manufacturer data in your region; OpenStreetMap Overpass query auto-populates with your coordinates once location is granted.
The observatory measures. The atlas narrates. Each chapter is one bioregion, read through whatever signal its people, its geography and its Fab City chose to put forward. One opening figure. Five hotspots. Every number earns a source. Synthetic where the connector isn't live yet, said plainly. This volume opens with Bali.
Volcanic ridge, river terraces, a port funnelling forty million tonnes a year — and a coastal village on Serangan island building its own green-hydrogen transition. Chapter one of the atlas.
The square that measured itself. How twenty residents with Smart Citizen Kits made Plaça del Sol quieter than any top-down noise study had managed — and why that story is the empirical floor PLANETAI stands on.
A country practising distributed survival. What a decade of economic collapse teaches a movement about resilience, sovereignty, and what "local" means when nothing else works.
What happens when a capital region, not just a city, decides to fab. Scale, policy, and the industrial question.
Before PLANETAI proposes that AI-coupled instrumentation can shorten the observation-to-response loop, the network already has four documented episodes where the loop closed at days-to-weeks cadence — three curated Fab City Challenge editions (Bali 2022, Bhutan 2023, Mexico 2024), and one network mobilisation under crisis (Mumbai 2020). Each note opens with its latency signal, holds three hotspots that earn it, and closes with what was learned. Sources are footnoted; challenge.fab.city carries the originals.
200+ villages onboarded to organic-waste exchange. 150+ recharge wells mapped. Five days from on-site kick-off to working prototype.
Asian Development Bank economist scored the portfolio at $5–12 per dollar spent. World Bank picked up the human–wildlife conflict project months later.
From Chiapas-jungle carpentry collectives to Monterrey biodiversity museums. Nine days from on-site kick-off to community evaluation in Cholula.
Maker's Asylum's M-19 collective shipped 1,073,704 face shields across 42 Indian cities and villages. Not an FCC edition — a network mobilisation under crisis.
Bali pledged in 2022 alongside four other Indonesian bioregions. The atlas reads this island through five signals: the volcanoes that make its soil, the port that feeds its kitchens, the aquifers holding its future, the rice that holds its culture, and the terraces that hold its slope.
Bali is a volcanic arc. Agung in the east, Batur in the centre — both active in living memory, both responsible for the soil that now grows four-fifths of the island's rice. The arable land is young. The rock is younger. Everything that feeds the island starts, literally, with a fault line.
Benoa handles roughly 7.6 million TEU per year — small on a world scale, outsized on an island one. A majority of the packaged food moving through the kitchens of Denpasar first crossed that dock. This is where the circular ambition of Fab City Indonesia meets the plainest fact of Balinese commerce: a lot of plastic flows in, not much flows out.
Ten kilometres up the coast from Benoa, Desa Serangan — 4,080 residents across six banjars and a Bugis village, 4.81 km² of mangroves, tidal flats, and reclaimed land — is the pilot site for Desa Hidrogen Hijau, the Green Hydrogen Village. The project, initiated by CAST Foundation, Meaningful Design Group and Fab Lab Bali with funding from ViriyaENB, runs the decarbonisation question backwards: instead of a national grid pushing renewables out to islands, a coastal community, on an island that lives off the sea, builds the electrolyser itself.
Indonesia still gets 86% of its electricity from fossil fuels. The state's Net Zero 2060 target is real, but rural and coastal villages — first to feel the climate crisis, last to be invited into the transition — rarely see the money. Serangan's response is fabrication-native: open-sourced, frugal, distributed, safe. A hydrogen fuel cell carries roughly 142 MJ/kg against a lithium battery's 1.8 — on an island with 4.8 kWh/m² of daily solar radiation and 633 km of coastline, that arithmetic starts to matter. This is what PLANETAI's Bali pilot means operationally: the Community tier measures an intervention that's already running.
The central aquifer under Tabanan has dropped measurably year over year. Groundwater abstraction by hotels on the south coast feeds the problem at the top of the hill. SYNTHETIC · CONNECTOR PENDING — exact drawdown figures pending the Indonesian Ministry of Water dataset, pending.
The subak water cooperatives are, in effect, a pre-industrial distributed manufacturing network for rice. A decision about water in one village downstream is made by committee upstream. A Fab City that doesn't learn from that is missing the point of being on this island.
Land-use change in the highlands is visible from 30 metres up, in any Landsat frame since 2014. Terraces abandoned for villas. Erosion onto the coast road. The question this chapter does not answer, because it can't: will the 2022 Fab City pledge translate into land-use policy before the next wet season does it for everyone.
Barcelona pledged Fab City in 2014 — the pledge that started the network. But the instrument PLANETAI stands on began earlier, in a small square in Gràcia, where twenty residents strapped Smart Citizen Kits to their balconies and turned a neighbourhood complaint into a measurement the city could not ignore.
Plaça del Sol sits in Gràcia — one of the livelier quarters of the old town, a neighbourhood that draws weekend crowds from across the city and beyond. For years, residents complained about nocturnal noise. The city heard it. Studies were commissioned. Nothing happened at the scale the plaza needed.
In 2016, as part of the EU H2020 Making Sense project, the neighbourhood association ran a focused eighteen-month campaign: twenty Smart Citizen Kits on balconies around the plaza, twenty residents, one question. Night-time readings regularly crossed 75–80 dB, well above the WHO's 40–55 dB residential guideline.
The square was redesigned. New paving, new seating, new acoustic treatment. Not because another top-down study demanded it — because residents had produced a measurement the city could no longer ignore. This is the campaign pattern PLANETAI's Community tier is built around: time-bounded, question-focused sensor deployments orchestrated by the local fab lab, not continuous monitoring fleets that decay before the data does.
Making Sense ran from 2015 to 2017. The Smart Citizen Kit it rode on began in 2012 at Fab Lab Barcelona and is still shipping — the v3 board is the campaign sensor of choice for fab labs across the network. The project is long over; the kit is now a fifteen-year campaign archive of European community-sensor data, deployed and retrieved as questions arose. That asymmetry is the Fab City thesis in miniature: campaigns end, fab labs outlast the politics that funded them — and the kit travels from one campaign to the next.
The superilla (superblock) programme that reshaped Barcelona's Eixample grid in the late 2010s was not a direct descendant of Making Sense. But it is continuous with the same proposition: that neighbourhood-scale environmental data, if governance reads it, is actionable — not in a five-year strategy, but in a street redesign. PLANETAI measures exactly this latency. That's what H₀-A is.
In November 2024 a DANA cut-off low dumped a season's rain on the Besòs watershed in hours. Google Flood Hub was live; its forecast now wires into the observatory's Bioregion panel. The Llobregat industrial corridor, the Zona Franca port, the Collserola ridge — these are the hotspots the next decade of Plaça-del-Sol-scale work will measure. Different resolution, same thesis.
Barcelona pledged Fab City in 2014. Twelve years in, the question this chapter leaves open is whether the city has been a net contributor to the movement it started, or a net beneficiary. The honest answer is both. Plaça del Sol is the strongest reminder that the movement's most important work still happens twenty balconies at a time.
October 2022. Fab City Foundation and Meaningful Design Group convened the first Fab Island Challenge on Bali, twinned with Bali Fab Fest (FAB22 + Fab City Summit). 92 participants from 23 nationalities. 200+ open-call applicants from 30+ countries. Eight areas of intervention drawn from the island itself — waste, water, food, materials, mobility, learning, ecosystems, value & exchange. Ten curated international teams were paired with ten local challenges and 32 challenge leaders, then put on the ground for ten days.
Empowering Sustainability with Plastic Exchange (USD 5K, public choice). The team wired together a buy-back model that has since rolled out across 200+ villages collecting waste — a proof-of-concept now operating well past Bali itself.
Supporting Smallholder Farmers with Kopernik (USD 5K, expert choice). The team built a market-access tool with Kopernik that supports 20+ farming families directly, with a network model designed to extend further.
Water for Life with IDEP Foundation (USD 2K Helium grant for LoRa). 150+ recharge / gravity wells were measured and mapped with low-power IoT — feeding into Bali's freshwater-table management. The same infrastructure pattern PLANETAI's Bali pilot will extend.
The total prize purse was USD 12,000 — a small budget for a large island. What it bought was access. The teams arrived with prototypes already shaped from twelve weeks of online co-creation; the on-site sprint converted them into working artefacts evaluated at Markerverse open day on October 21–22, with 256 public and expert votes cast across ten teams.
The honesty marker: this was not an instrumented experiment. There is no synthetic-control district, no pre-registered baseline, no quantified counterfactual against which the 200-village figure was measured. What it is is a documented operational cadence — brief published, applicants converged, prototypes shipped, communities adopted — at a tempo no municipal procurement cycle in Bali has ever matched.
PLANETAI's Bali pilot picks up on the same ground. Fab Lab Bali, IT Del, CAST, MDG. The Tri Hita Karana ↔ four-pillar mapping is run as a Bali pilot only co-design experiment with a named Balinese customary authority; publication is conditional on customary-authority sign-off. The instrument PLANETAI proposes is what would let the 2027 cohort report a pre-registered, district-level ρ_community against this 2022 baseline.
The brief→prototype cadence is bounded — between 12-week online build-up and a 5–7-day on-site sprint, the action layer is operational. The downstream uplift figures (200+ villages, 20+ farming families, 150+ wells) are documented but not pre-registered or counterfactual-controlled — they belong to the priors PLANETAI is built to convert into a measurement, not to the measurement itself.
July 2023. Fab Bhutan Challenge ran inside FAB23 in Thimphu, hosted by DHI Innotech and the Jigme Namgyel Wangchuck Super Fab Lab, with Fab Foundation, MIT Center for Bits & Atoms, and Fab City Foundation. Five host fab labs across Bhutan, 59 participants from 22 countries, 250+ students engaged, 15+ community partners, twelve weeks of online co-creation and a seven-day on-site sprint. Five challenges: climate-adaptive agriculture, water conservation, human–wildlife conflict, culture preservation, assistive technology.
Asian Development Bank economist Milan Thomas applied a Kremer-style innovation-portfolio valuation to the five challenges, returning a base-case BCR of 8.24 with a $5–12 sensitivity range. Even if only the lowest-return solution succeeds, BCR ≈ $2 — the challenge still pays for itself twice over. The author flags every estimate as a conservative lower bound.
Months after the Challenge, the Human-Animal Conflict project received World Bank funding for regional deployment. This is the strongest realised post-edition outcome across all three FCC editions, and the most defensible H₀-A-relevant cadence signal in the network's record.
Aluminium Waste, Gracefully Braced (Fab Lab CST) shipped 8 distinct assistive-tech projects for special-educational-needs students plus a 5-year STEAM education plan for the College of Science and Technology. Weaving Technology into Tradition (JNW Super Fab Lab) put an enhanced loom and an online repository of weaving patterns into the hands of 200+ Bhutanese students. USD 10K total seed.
The Bhutan edition is the only one in the network's record with an independent, peer-presented economic evaluation. Thomas (2023), an ADB economist working out of the Bhutan country office, used the 2022 Bhutan Labor Force Survey, the 2017 Population & Housing Census, and the 2022 BLSS to build conservative reach assumptions per challenge — only 20% of 2,800 Thimphu weavers reachable, 50% of Sarpang's 6,000 farmers — then weighted each by an explicit elicited probability of success (50% / 80% / 70% / 75% / 90%) and discounted to 2023 USD.
The headline number that survived all that conservatism: BCR 8.24, social rate of return 145% through 2033. Total benefits $1.97M against costs of $239K. The author is explicit that he excluded cultural-preservation value, omitted human and animal lives saved from the wildlife project, omitted public-health value from clean water, and omitted cross-region scaling. The lower bound is what came back at $5–12.
For PLANETAI, the Bhutan edition does two things at once. It establishes that the network can produce a deployable prototype in twelve weeks of online build plus seven days on the ground, and that an independent economist with no Fab City affiliation will score the portfolio at multiples of cost. The H₀-A null we will test is whether AI-coupled instrumentation moves ρ against this baseline — and whether a quantified BCA of similar discipline can be produced for Barcelona, Boston, Santiago, and Bali.
An independent ex-ante BCA exists for the network and reads positive at conservative lower bounds — but it is projected, not realised. Realised outcomes (World Bank pickup) are documented but not synthetic-control measured. PLANETAI's contribution is to take the projection method and replace its probability-of-success priors with measurements.
July–August 2024. Fab City Challenge Mexico ran inside FAB24 in Puebla, hosted by Ibero Puebla's IDIT with Fab Foundation, MIT CBA, and Fab City Foundation. 36 challenge participants from 12 countries split across seven host locations from Monterrey to Cancún. Six areas of intervention: biodiverse cities, water conservation, social entrepreneurship, accessible assistive tech, bio practices, regenerating neighbourhood perspectives. The cohort sat inside a wider FAB24 audience of 1,200+ attendees from 55 countries.
BiodiverCities — Mobio (Innovaction Gym + LABNL, Monterrey). 500+ public votes. A modular mini-museum on biodiversity that tours and folds back, designed to deploy at neighbourhood scale rather than wait for a permanent venue. USD 5K seed.
Crafting in the Jungle (IXIM A.C., San Jerónimo Tulijá, Chiapas). A carpentry and furniture-design empowerment programme co-designed with the women of Yo Tan Te. The team set up a CNC-supported workflow that the collective can run without the visiting team — economic independence on local terms. USD 5K seed.
Three lasting outputs from one neighbourhood — a FabLab Empowerment Programme, an Art Toy cottage industry, and the Our VOP media-activism app — alongside an urban-furniture community-space revitalisation with FARO. The network's clearest example of a single team leaving a stack behind.
The Mexico edition extended the format. Seven host cities instead of one or five. The longest pre-event build of any edition (16+ weeks). The largest open-day audience (2,000+ visitors at CIEN, Cholula, on August 3–4). 85 open-call applicants from 23 countries, narrowed to 42 curated participants who were then deployed to host fab labs across the country.
Honesty marker: the publication's impact section is narrative-only — no quantified beneficiary numbers are published. The two pages of impact prose describe "a catalyst for transformative change" without N-figures. If a future evaluation needs a beneficiary count, it must be sourced from internal FCF records, not from the public retrospective. This is a known gap, and one PLANETAI-style instrumentation would close.
What the edition does prove is portability. The same 16-week-online + 9-day-on-site cadence holds across a different continent, a different statistical system, a different language stack, a different host topology. The Bali pattern was not Bali-specific. That portability is the precondition for measuring ρ across PLANETAI's four bioregions on the same scale.
The cadence is portable across geographies. The downstream impact is not yet quantified at the Bhutan-edition level of rigour. PLANETAI takes the same workflow and adds the instrumentation that produces a measurement, not just a publication.
March 2020 onward. India's COVID-19 wave. Maker's Asylum's M-19 collective — a fab-lab-rooted volunteer network — moved from a 3D-printed prototype to a laser-cut sheet workflow once the demand curve made the 3D-print path untenable, then shipped face shields across 42 Indian cities and villages over 49 days. Not a Fab City Challenge edition. A network mobilisation under crisis. The single strongest action-latency signal in the literature PLANETAI sits inside.
M-19 began with 3D-printed shield frames; demand outpaced print speed within days. The collective's pivot to laser-cut polymer sheets — a different process, different material, different tolerance budget — was made and propagated across the volunteer network without a central design authority slowing it down.
The peer-reviewed cross-case literature [Corsini et al. 2021; Kieslinger et al. 2021; Tabarés et al. 2021; Hamdi-Cherif et al. 2021] reads M-19, Isinnova in Brescia, Coronavirus Makers in Spain, and the Tunisian Solidarity FabLabs as one network — innovation communities whose response speed was infrastructure-built, not crisis-built.
No study in this body of work measures action-latency reduction against a controlled policy-only counterfactual. What the literature establishes is detection-to-prototype and detection-to-territory-deployment speed, for low-regulatory-burden artefacts, under crisis conditions. That is exactly the gap H₀-A is built to fill in non-crisis steady state.
M-19 is the loudest of a quiet pattern. Across COVID, the global fab-lab network produced documented detection-to-prototype cycles of hours — Isinnova in Brescia reverse-engineered a missing Venturi-valve prototype within six hours of a hospital running out of supplier stock on 13 March 2020, with 100 working units in 24 hours and patients on the device the following day. Detection-to-territory cycles of weeks — Barcelona's Ateneus de Fabricació, declared an essential service in late March, met PPE demand for all hospitals in the city and metropolitan area by the third week of April.
Outside crisis time, the same pattern shows up at lower stakes. e-NABLE has delivered 10,000–15,000 upper-limb prosthetics through ~40,000 volunteers across 100+ countries. Maker's Asylum is the same network, the same workflow logic, the same trust topology — just under different load.
PLANETAI is built on a hypothesis about that infrastructure: that AI-coupled instrumentation can shorten the observation-to-response loop in steady-state conditions, the way M-19 did under crisis. The instrument is what lets us test it against a pre-registered policy-only baseline rather than just narrate it.
The action layer can move at hours-to-weeks cadence under load. The literature documents speed; it does not measure speed against a controlled counterfactual. PLANETAI's H₀-A is the experiment that would convert the M-19 pattern into a falsifiable claim.