Each showcase asks one question of public data. Public tables often name a school, a sensor or a building but carry no shape, so they cannot be measured or mapped. Each page says what was missing, how the shapes were made in Plotlay and keyed to the data, and what the numbers show. Every one has a map you can open without an account, and its data, Python notebook and Power BI files on GitHub.
Inner Sydney and Parramatta · NSW open data
- The analysis
- Land per student at 81 government schools, 2015 to 2025.
- The problem
- The department publishes every school’s headcount, but not the land each school sits on.
- In Plotlay
- Each school outlined from NSW Cadastre lots, keyed by school code, the headcounts joined and coloured by m² per student on a time slider.
Read the showcase →GitHub
Inner Sydney and inner Adelaide · NSW and SA open data
- The analysis
- Spending, outcomes and school grounds in two states: each on its own, then side by side.
- The problem
- Neither state publishes how much land its schools sit on, and Adelaide has no open cadastre or open aerial photos.
- In Plotlay
- Outlines from NSW lots in Sydney and from the state’s open land use file in Adelaide, then one method for both cities.
Read the showcase →GitHub
Surry Hills, Sydney · NSW open data
- The analysis
- Why a 14-storey school is the exception in a land analysis, and its floor space per student, estimated, over its staged opening.
- The problem
- Land per student says how a tower was built, not how much room its students have, and no open dataset gives the floor area.
- In Plotlay
- The site from cadastre lots, each building picked or drawn with its storeys, floor space as a formula column, divided by each year’s headcount on the time slider.
Read the showcase →GitHub
Parkville, Melbourne · City of Melbourne and Victorian open data
- The analysis
- How much room one school has, building by building: grounds and floor space (estimated) per student, then a facilities section in sample data: assets, risk, work orders, meters.
- The problem
- School maps draw one shape per school, and the council’s land use census has 2 records for 13 buildings.
- In Plotlay
- Each building picked from the council’s footprints, floor space estimated from its height, then rolled back up to the site.
Read the showcase →GitHub
Melbourne CBD · City of Melbourne open data
- The analysis
- How crowded 53 city footpaths are, hour by hour, on weekdays and weekends.
- The problem
- The council publishes hourly counts from its sensors, but a count alone does not say how much footpath each person has.
- In Plotlay
- A footpath zone drawn for each sensor, keyed to its counts, and coloured by people an hour on each metre of width through the day.
Read the showcase →GitHub
City of Melbourne · City of Melbourne open data
- The analysis
- Where a council uses its electricity, gas and water, from its own bills.
- The problem
- 5,444 bill rows name an asset, but 106 of the council’s 316 assets have no location in its asset register.
- In Plotlay
- The missing assets placed from the council’s parcels, footprints and streets, bills spread by day and coloured by use.
Read the showcase →GitHub
City of Melbourne · City of Melbourne open data
- The analysis
- Which streets and blocks lose their council-tree shade over the next 20 years.
- The problem
- The council lists 82,064 trees with their years left, and maps its canopy as outlines, but neither says which streets or blocks lose what.
- In Plotlay
- All 603 census blocks taken from the council’s file in one step, keyed by block_id, the trees placed in them by position, and the decades on a slider.
Read the showcase →GitHub
Melbourne CBD · City of Melbourne open data
- The analysis
- Where office floor space gave way to apartments, hotels and student rooms, block by block, 2002 to 2024.
- The problem
- The council’s land use census is a table keyed by block, with no shapes, and it holds back small figures.
- In Plotlay
- All 603 blocks keyed by block_id, the census linked from the portal with the year on a slider, formula columns for the change, charts made in the Analysis workspace.
Read the showcase →GitHub
Melbourne CBD · City of Melbourne open data
- The analysis
- How full the kerb is by hour and street, weekdays against weekends, and how long cars stay.
- The problem
- The council’s parking sensors log 42.7 million arrivals and departures that name a street but have no outline to map.
- In Plotlay
- The council’s road corridors keyed by seg_id, each street’s summary and hours linked, an hour slider, charts made in the Analysis workspace.
Read the showcase →GitHub
New York City · NYC Open Data
- The analysis
- New York City’s Local Law 97 (2019) caps the carbon emissions of most buildings over 25,000 square feet, with a fine for every tonne over and caps that tighten in 2030. Each benchmarked building’s 2024 emissions against its limit, and the likely penalty, now and from 2030.
- The problem
- The city publishes 39,090 records of energy use and floor area, but not each building’s limit, its emissions as the law counts them, or a shape.
- In Plotlay
- Every district and Midtown footprint taken at once and keyed by its id, every property placed by its position, coloured by period.
Read the showcase →GitHub
San Francisco · DataSF and California open data
- The analysis
- San Francisco requires older wood-frame apartment buildings with a weak, open ground floor (garages or shops) to be strengthened against earthquakes. Every building on that program’s list, by the year its retrofit was finished, against the state’s mapped liquefaction zones.
- The problem
- The list gives each building’s status today, but not when it was finished or whether it stands on ground that can liquefy.
- In Plotlay
- 4,901 parcels taken at once and keyed by block and lot, each permit’s year as a time axis, the zones over them.
Read the showcase →GitHub
United States · National Park Service open data
- The analysis
- Visits per km² of each park’s visitor core against the whole park, month by month since 2015, for the 20 most visited national parks and five contrasts.
- The problem
- The Park Service counts visits per park and month, with no place in the count; spread over a whole park, the crowds disappear.
- In Plotlay
- 25 parks and 25 cores drawn by one rule over the USGS aerial photo, keyed by unit code, 132 months on the slider.
Read the showcase →GitHub
New York City · NYC Open Data
- The analysis
- The city’s 1995, 2005 and 2015 street tree censuses compared neighbourhood by neighbourhood: +35% more live street trees.
- The problem
- Each census is its own table, with its own columns and spellings, and none compares the three.
- In Plotlay
- Every tree placed by its position into its neighbourhood, the censuses on a slider, East New York tree by tree.
Read the showcase →GitHub
Plotlay
