# urban:SCAN > Urban analysis of any place on Earth from open data, in the browser. urban:SCAN builds a 3D model of a site (buildings with heights, ages and uses, streets, transit, trees, water, terrain, imagery) and runs 32 analyses of form, use, movement and environment on it, with a graphics editor and exports to DXF, GeoTIFF, GeoJSON, CSV, PNG, GLB, KMZ, a vector PDF report and video. Made by IL DARK (https://ildar.space). - App: https://urban-scan.ildar.space/app - Landing page with the full catalogue: https://urban-scan.ildar.space/ - Pricing: https://urban-scan.ildar.space/pricing ## How it works 1. Territory: search a place, choose the site scale (up to 20 km²) or the city scale (up to 20 km across), a square or a circle, or upload a boundary (KMZ, KML, GeoJSON). 2. Analysis: open any analysis and tune it; everything runs in the browser (Web Workers, WebGL) and is free to look at. 3. Graphics (optional): colours, gradients, line widths, labels, legends (wind rose, sun ruler, sun path) of the plan and the exploded 3D scheme. 4. Download: DXF, GeoTIFF, GeoJSON, CSV, PNG, GLB, KMZ, PDF report, video. Pricing: free to explore; downloads cost credits (10 for $5, 30 for $12, 100 for $30, 200 for $50). One analysis on one site is paid once and downloads in every format. Simple analyses 1 credit, network and visibility 2, simulations 3, wind and rain 5, PDF report 8; sites over 5 km² cost in proportion to the area. ## Analyses ### Form - **Building heights** (1 credit): Height and number of floors of every building, where the height comes from, and the footprint source. - How: Each footprint takes the best height available: an OSM height tag, OSM floors × 3.2 m, Overture or Microsoft heights, then the WSF 3D building height estimate of the 90 m cell, then a typical height for the building type. Views: floors, height in metres, height source, footprint source. - Data: OpenStreetMap, Overture Maps (Microsoft, Google and East Asian footprints), Microsoft ML Building Footprints, WSF 3D v2 (DLR). - **Building age** (1 credit): Year of construction of the buildings by epoch, and the growth of the built-up area from 1985 to 2015. - How: Exact years come from OSM start_date tags, open cadastres and Wikidata. Elsewhere the first year the land appeared built up in WSF Evolution bounds the building from below. The map is offered where exact years cover at least 60% of the built-up area. - Data: OpenStreetMap, cadastres of New York (PLUTO), Philadelphia, Chicago, Calgary, the Netherlands (BAG), France, Switzerland, Vienna, how-old-is-this.house (80 Russian cities), Wikidata, WSF Evolution (DLR, Landsat). - **Built density** (1 credit): Floor area ratio, ground coverage, dwellings, residents or jobs per hectare as a smooth surface. - How: Gross floor area (footprint × floors) of every building is spread with a Gaussian moving window of a chosen radius. Dwellings come from NYC PLUTO where available, else residents / 2.4. - Data: Building footprints and heights of the site (see Building heights), NYC MapPLUTO. - **Blocks and Spacematrix** (1 credit): Urban blocks between streets classified by Spacematrix: low-rise, mid-rise or high-rise, compact or open. - How: Blocks are cut out by the street network. For every block FSI (floor space index), GSI (ground coverage) and L (number of layers = FSI / GSI) are computed and placed on the Spacematrix of Berghauser Pont & Haupt. - Data: OpenStreetMap streets, building footprints and heights. - **Street canyons** (1 credit): Height-to-width ratio of streets: how enclosed, shaded and wind-sheltered each street is. - How: Every 12 m along a street, rays find the first buildings on both sides; their heights divided by the distance between them give H/W. - Data: OpenStreetMap streets, building footprints and heights. - **Visibility and sky view** (2 credits): How much space a pedestrian sees (isovist area) and how much of the sky is open (sky view factor). - How: Isovists are cast at 1.6 m from every open cell with 48 rays up to 250 m over buildings and tree crowns. The sky view factor is the share of the sky above the horizon in 16 directions within 120 m. - Data: Building volumes, tree crowns, terrain model. - **View corridors** (2 credits): Where a landmark, water or green can be seen from the street. - How: From every open cell at 1.6 m, view directions are traced up to 600 m; the share of directions that reach the target is mapped. - Data: Building volumes, tree crowns, terrain, OSM water, parks and landmarks. - **Cadastral parcels** (1 credit): Land parcels with their size, site coverage and floor area ratio. - How: Parcels are downloaded from open cadastral services; each building joins the parcel that holds its centre. Coverage = footprint / parcel area, FAR = floor area / parcel area. - Data: Kadaster (Netherlands, PDOK), IGN Parcellaire Express (France), ALKIS (Berlin), Maa-amet (Estonia), Philadelphia Water Department parcels. - **Satellite** (1 credit): Aerial or satellite image of the site and vegetation, water and built-up indices. - How: National orthophotos where they exist, Sentinel-2 elsewhere. Indices (NDVI, EVI, SAVI, NDWI, NDBI and others) are computed from Sentinel-2 bands in the browser. - Data: National aerial imagery, Sentinel-2 cloudless (EOX), Sentinel-2 L2A. ### Use and people - **Land use program** (1 credit): Use of every building (housing, offices, retail, industry, public) and the mix of uses around. - How: Uses come from building tags, then from the places and land use inside the footprint. The mix is the Shannon entropy of floor area by use within about 150 m (0 — one use, 1 — an even mix). - Data: OpenStreetMap buildings and land use, Overture places. - **Activity points** (1 credit): Shops, cafés, services and other places, their density and diversity. - How: Places are merged from OSM and Overture, deduplicated and classified. Density per hectare and the number of different kinds are smoothed with a chosen radius. - Data: OpenStreetMap points of interest, Overture places. - **Residents and jobs** (1 credit): Estimated residents and jobs per building and per hectare. - How: Residential floor area × 0.78 divided by living space per person (set for the country); small houses hold one household. Jobs follow floor area by use. At the city scale WSF 3D volume and the residential share are used. - Data: Building floor areas and uses, WSF 3D (DLR). - **Housing value** (2 credits): Where housing is likely dearer: a relative location value of homes, 0–100, or approximate prices per m². - How: An index of access, transit, amenities, parks, water, quiet, newness and height. With a city average price the index becomes approximate prices. In New York official assessed values are used directly. - Data: All site layers; NYC MapPLUTO assessments. An estimate, not observed prices. - **15-minute city** (2 credits): Walking time to everyday services and how many of them are within reach. - How: Network walking time at 4.8 km/h from every cell to groceries, food, health, education, park, stop, shops, services, sport and culture. - Data: OpenStreetMap streets and paths, places from OSM and Overture. - **Public spaces** (2 credits): Squares, pedestrian streets, parks, playgrounds and markets scored for how good they are as places. - How: Score 0–100 after Gehl: active ground floors 25%, people within 5 minutes 20%, green 15%, benches and furniture 15%, enclosure (best H/W 0.3–0.8) 15%, entrances 10%. A type view shows what kind of space each one is. - Data: OpenStreetMap, Overture places, buildings and residents of the site. - **Walkability** (2 credits): Where walking is easy: dense, connected, mixed, with services and transit nearby. - How: Walkability index after Frank et al. (2010) — residential density, intersection density, land use mix — extended with services within 15 minutes and transit. - Data: Streets, buildings, residents and places of the site. ### Movement - **Space syntax** (2 credits): Through-movement and to-movement potential of every street: main streets, centres and back streets. - How: Angular segment analysis of the street network around the site (Turner 2007, Hillier et al. 2012): normalised choice (NACH), normalised integration (NAIN), choice and mean depth at radii from 400 m to the whole network. - Data: OpenStreetMap street network (walking or driving). - **Pedestrian flows** (3 credits): Estimated people per hour on every street at a chosen time of day, with an animation of walkers. - How: Trips from homes, jobs, transit stops and places, destinations chosen by a gravity model weighted by the time of day, routed along shortest walking paths. - Data: Streets, residents, jobs, stops and places of the site. - **Car traffic** (3 credits): Vehicles per hour and congestion (volume to capacity) at the peak hour. - How: Gravity trip model with through traffic entering at the main roads, assigned with congestion (BPR function). Line width is vehicles per hour, colour is V/C. - Data: OpenStreetMap roads (class, lanes, speeds), residents and jobs of the site. - **Public transport** (1 credit): Stops, routes with their colours, service frequency along streets and walking distance to a stop. - How: Route relations and stops are read from OSM; walking distance is measured along the network; rail counts double. - Data: OpenStreetMap public transport routes and stops. - **Cycling** (1 credit): Low-stress cycling network and the residents who live near it. - How: Level of traffic stress (LTS 1–4) after Mekuria, Furth & Nixon (2012) from road class, speed and cycleway tags; residents within 250 m of the LTS 1–2 network; bike parking and rental. - Data: OpenStreetMap cycleways and cycleway:* tags. - **Isochrones** (2 credits): How far you get in 5, 10, 15… minutes from a point or an area, and what each band reaches. - How: Travel time along the street network: walking 4.8 km/h, cycling 15 km/h, driving at street speeds. - Data: OpenStreetMap street network; residents and places of the site for the counts. - **Parking** (1 credit): Parking lots, garages and street parking with their capacity, against the cars of the residents. - How: Capacity from OSM or about 25 m² per space; street parking from parking:* tags; demand from residents and car ownership per 1000 people. - Data: OpenStreetMap parking. ### Environment - **Insolation and shadows** (3 credits): Sun hours on the ground and on facades on a chosen date, real-time shadows and the sun path. - How: Sun positions every 10 minutes; buildings, terrain and tree crowns cast shadows on the GPU. Facade points at 1.6 m and mid-height; buildings show the share of facade points that meet the norm. Legend: sun ruler and sun-path diagram. - Data: Building volumes, terrain, tree crowns; solar position (NOAA algorithm). - **Pedestrian wind** (5 credits): Wind speed at pedestrian level relative to the free wind, and comfort classes. - How: A lattice Boltzmann simulation on the GPU, 2D slice at pedestrian level, for the prevailing or a chosen direction (eight directions for comfort). Comfort by Lawson (LDDC). Legend: wind rose of the year. An estimate, not CFD. - Data: ERA5 reanalysis (ECMWF) hourly wind at 10 m via Open-Meteo, building footprints and heights. - **Traffic noise** (3 credits): Day noise level from roads, trams and rail in dB. - How: RLS-90 emission from the traffic model, rays with screening by buildings and air absorption; trams and rail by typical service. - Data: OpenStreetMap roads and rail, the car traffic model, buildings. - **Rain and puddles** (5 credits): Where water runs and stands during a storm and the hour after it, as an animation. - How: Fill-and-spill of the terrain hollows with runoff particles; carriageways lie a kerb lower, roofs drain into the sewer, lawns soak 25 mm/h, drains take the given capacity. - Data: Terrain model (Mapterhorn, Copernicus GLO-30), streets, buildings and green of the site. - **Trees and 3-30-300** (1 credit): Tree canopy and the 3-30-300 rule: 3 trees seen from home, 30% canopy around, a park within 300 m. - How: Trees from OSM and crowns detected on aerial imagery or the canopy height map; checks per home after Konijnendijk (2021). Species lists from iNaturalist and GBIF. - Data: OpenStreetMap trees and parks, national aerial imagery, Meta / WRI Global Canopy Height, ESA WorldCover. - **Heat island** (2 credits): Land surface temperature on a hot clear day and its anomaly against the surroundings. - How: Surface temperature of a clear summer Landsat scene in °C; the anomaly is the difference from the median temperature of tree, shrub and grass cover on the site. - Data: Landsat 8/9 Collection 2 Level 2 surface temperature (Planetary Computer), ESA WorldCover 2021. - **Water flow and TWI** (2 credits): Where water runs over the ground, the topographic wetness index and the depressions where water would stand. - How: Depressions filled (priority flood), flow by steepest descent (D8) and spread downhill (MFD); channels where the upslope area exceeds a threshold; TWI = ln(a / tan β). Animated flow particles. - Data: Terrain model: lidar where Mapterhorn has it, Copernicus GLO-30 elsewhere. - **Risks and constraints** (1 credit): Official flood zones, protected areas and heritage on the site, with the share of the site in each. - How: The layers are read from open services for the site and laid on its grid; a cell takes the most binding class (heritage, then the 1% flood zone, the 0.2% zone, protected areas). The passport turns the shares of the plot into risks and constraints. - Data: FEMA National Flood Hazard Layer (USA), WDPA protected areas (UNEP-WCMC, IUCN, worldwide), OpenStreetMap heritage, historic and listed objects. - **Solar roofs** (3 credits): Annual sunlight on every roof and the power a PV system could produce. - How: Annual global irradiation, the direct part reduced by shadows of buildings, terrain and trees (a day per month on the GPU); PV on 60% of the roof, 21% modules, 82% performance ratio. - Data: ERA5 irradiation via Open-Meteo, building volumes, terrain, trees. ## Data sources - Streets, buildings, places, land use: OpenStreetMap contributors (ODbL), Overture Maps Foundation - Building footprints: Microsoft ML Building Footprints, Google Open Buildings and East Asian buildings via Overture - Building heights and age: WSF 3D and WSF Evolution (DLR), open cadastres, Wikidata - Terrain: Mapterhorn (lidar in ~70 countries), Copernicus GLO-30 DEM - Imagery: National orthophotos, Sentinel-2 cloudless (EOX), Sentinel-2 L2A, Landsat 8/9 - Climate: ERA5 reanalysis (ECMWF) via Open-Meteo - Trees and land cover: Meta / WRI Global Canopy Height, ESA WorldCover, iNaturalist, GBIF - Cadastral parcels: Kadaster (PDOK), IGN, ALKIS Berlin, Maa-amet, City of Philadelphia - Map: OpenFreeMap, OpenMapTiles ## FAQ **What is urban:SCAN?** A web tool that builds a 3D model of any place on Earth from open data and runs 32 urban analyses on it in the browser — from space syntax and density to insolation, wind, noise, heat and water flow — with a graphics editor and exports to CAD, GIS and a PDF report. **Which places does it cover?** Any place on Earth. Streets, buildings and places come from OpenStreetMap and Overture, which cover the whole world; heights, ages and terrain have global fallbacks (WSF 3D, WSF Evolution, Copernicus GLO-30). Some layers are richer where open data is: lidar terrain in about 70 countries, exact building years from open cadastres, cadastral parcels in the Netherlands, France, Berlin, Estonia and Philadelphia. **How large can a site be?** A site of up to 20 km² (a 4.47 km square or a 5 km circle) with every analysis in full detail, or the city scale — a coarse analysis of a whole city up to 20 km across. **Do I need to install anything or sign in?** No. The app runs in a modern browser with WebGL. You only sign in with an email link when you want to download files. **How accurate are the simulations?** They are fast estimates for early design and research: insolation is computed exactly from the sun position and the 3D model; wind is a 2D lattice Boltzmann slice at pedestrian level, not CFD; flows and noise come from simplified transport and acoustic models; housing value is a relative index, not observed prices. Every analysis states its method and data. **What is space syntax and which measures are there?** Space syntax describes how the configuration of the street network shapes movement. urban:SCAN runs angular segment analysis and shows normalised choice (NACH, through-movement), normalised integration (NAIN, to-movement), choice and mean depth at radii from 400 m to the whole network. **Which file formats can I download?** AutoCAD DXF with layers, GeoTIFF, GeoJSON, CSV of buildings, PNG, a GLB 3D model, Google Earth KMZ, a layered vector PDF report (editable in Illustrator) and video of the 3D view or the scheme. **How much does it cost?** Looking at everything is free. Downloads cost credits: 10 for $5, 30 for $12, 100 for $30, 200 for $50. One analysis on one site is paid once and then downloads in every format; PNG pictures are free on sites up to 5 km². **Can I test a project?** Yes: draw new volumes with floors and a use, or remove existing buildings, and compare any analysis before and after. Your own KMZ, KML, GeoJSON and GeoTIFF layers can be laid over the site and polygons can become project volumes. **Can I add the GIS layers of my city?** Yes. Paste the address of any open WFS service or ArcGIS feature service — zoning, cadastre, heritage, flood maps of your city — and pick a layer: the features of the site come in with all their attributes, show in the 3D view, the map, the scheme and the DXF, and their attributes open with a click. KMZ, KML, GeoJSON and GeoTIFF files work too. **Can I share a site with a client or a team?** Save the site as a project and share its link: anyone can open it without an account and recompute it in the browser. Signed-in people can leave comments pinned to places on the map. Saved projects can be compared side by side by their site passports. **What is the site passport?** Draw or upload the boundary of your plot inside the study area: the passport puts the numbers of the plot next to those of the whole area — floor area, FAR, coverage, heights, residents, green, the mean of every opened analysis, flood zones, protected areas and heritage — and turns them into strengths, constraints and risks, with the thresholds written out. It comes as CSV and as a page of the PDF report. **Is the code open source?** No. The data it uses is open and credited; the tool itself is a commercial service by IL DARK.