Australian Remote Operations for Space and Earth
Landscape mapping · v26 dataset · 4 November 2026
Australian STEM Engagement Map
This is the most recent version of the map, updated with industry and community feedback. The "Australia's STEM Ecosystem, Mapped" Report references the v19 dataset which can be viewed here. This latest update increased the program count, fixed some issues with program details, adjusted the VET program data, and removed entries that were found to have ended.
This displays every mapped initiative placed against the learning continuum. Cells and bars are entries in the ecosystem map, not programs; grouped institution types count once. 287 of the 308 entries link straight through to the provider — click any bar in “how far each initiative reaches”.
Where sustained contact actually happens
How many structured contacts are available, with life stage across and engagement dosage down.
Vocational programs are not included in this, as dosage has not been quantified.
How far each initiative reaches
One bar per entry, visualising the life stages reached and the type of delivery.
Click on any name to open the provider's website.
How to read it. Pale segments are partial or delivery-model-limited coverage (the half-filled marker in the workbook). Long bars are rare: most entries touch one or two adjacent stages. Click any name to open the provider’s site. The 21 that do not are the grouped category rows (all botanic gardens, all universities, the competition and utility aggregates), which stand for many providers and so have no single address.
Where it is available
Most entries are national, so counting what is available in each jurisdiction separates them barely at all. What does separate them is how many are based in each.
What the two numbers mean. Available is what can be reached from that jurisdiction and it barely separates them, because the national layer lands everywhere. Built here counts entries that run in that jurisdiction and nowhere else. That is the capacity to originate something locally, should be read as capacity not performance. A small jurisdiction with few local initiatives is not failing, it is small.
Read “built here” as what data was available, not how much each jurisdiction does. Five jurisdictions have had a state-wide source folded in, and they are not equivalent. Queensland, Tasmania, Western Australia and South Australia have current datasets, and Victoria one from 2017. NSW, NT and ACT had no findable directory for this project to sweep, and data is subsequently skewed.
Solid is substantive presence; pale is partial or limited by delivery model. The rural and remote bar is the one to watch — a large share of its length is partial coverage, much of it digital reach rather than presence on the ground.
Why there is no dot map. Three quarters of the geographic coding is inferred from reach and delivery model rather than verified, and a pin map would show head offices rather than who is reached. These are stated availability, not delivery volume: an entry counts once in a jurisdiction whether it runs one workshop there or two hundred.
A note on the partial marker. At jurisdiction level it is barely used — three to five entries per state, about 2% — so availability there is effectively yes or no. The substantive-versus-partial texture lives in these setting bars instead, where regional carries 56 partial marks and rural and remote 62.
The entries
Every value in the panels above, readable as text. Click a column to sort.