A day of desktop constraint screening, in minutes.
SiteCensus queries 450+ live government layers against your site in a single run, resolves them into structured findings, and produces the assessment document — with the reasoning and the full record behind it.
Queensland from 31 August 2026 · New South Wales next · No install, no GIS licence
The data is authoritative and public. Assembling it into an answer is where the time and the risk sit, and the bill arrives more than once per project.
Commonwealth and State instruments sit on separate services run by separate agencies, under different names for overlapping matters. A dozen portals, worked by hand.
Move the boundary, add an alternative alignment or widen the scope and the entire search repeats. The cost is not one day — it is one day per iteration.
Two people screening the same site produce different results, so a reviewer redoes the search rather than checking the reasoning. Senior time is spent twice.
A dataset nobody opened appears once the design is fixed and the application lodged, when resolving it costs a redesign rather than an afternoon.
One workspace that queries authoritative data live, resolves it into structured findings, and produces the document that goes on file.
Establish the study area however suits the job, choose what to check, and run a single intersect across every enabled layer whose data could reach the site. Everything is off by default, so a run reflects a deliberate scope.
Overlaps are dissolved so one constraint is counted once. Findings group into five categories with area, percentage of site, and separate site and buffer figures. Each is matched to the instrument it triggers, the pathway and the administering agency.
A branded report, in full or tailored through a short project intake, carrying the constraint rating and the rule behind it, the trigger matrix, extent of overlap, conservation-significant species, recommendations, limitations and the complete layer register.
Terrain from national elevation data, four decades of satellite record, and a print composer that produces the figure for the report instead of a screenshot.
All layers sourced live from government and other authoritative services. Update frequency varies by custodian and is recorded per layer in the application.
The same structured output answers a different question depending on who is holding it.
Every new site needs the same desktop constraint picture before anything else can proceed. Specialist time goes into interpretation rather than retrieval, and a reviewer checks the reasoning instead of repeating the search.
Route and layout decisions lock in early. Screening at concept stage turns a trigger into a design input rather than a redesign, a resubmission and a schedule impact. Corridors screen section by section.
The same constraint picture across every project, from the same live data, produced the same way each time. An officer can screen a site a consultant has written about and compare, without commissioning further work.
A screen is only useful if the output can be checked. Four things are fixed by design.
Every layer is fetched from the custodian's REST service at the moment the assessment runs. Currency is shown per layer in the application.
A layer that fails its query is reported as failed rather than returned as zero. A service that cannot be reached is named, with its subject matter marked not covered. A run with no layers selected will not proceed.
Printed in the report with the matching row marked, and not adjustable from it. Modelled evidence never raises the rating on its own.
Layers queried, layers returning features, distinct constraints and total features are stated with the reasons they differ. A layer exposed twice is counted once.
Stated here, and restated in every report.
Desktop screening, not field validation. It does not assess condition or significance on the ground.
A spatial intersect is binary. It establishes that a constraint is present, not how significant it is on your site.
Not legal advice. Statutory triggers and timeframes are indicative and must be verified with the administering agency.
A finding of no intersection is not a clearance under any legislation.
RE and regulated vegetation mapping is broad scale and indicative at property scale. Ground-truthing or a PMAV is required before clearing decisions.
Species records are presence-only and historical, precision up to 10 km. Absence of a record is not evidence of absence.
Uploaded and connected layers are reference-only and are not screened.
Queensland launches 31 August 2026, New South Wales follows. The platform is in final testing.
Each state version is built against that jurisdiction's own legislation, datasets and agencies, which is why they are released one at a time. Every subscription includes each new state as it launches, at no additional cost.
All prices in AUD, GST added at checkout. Every plan includes the full platform and every new state as it launches.
1 user
per month · or $1,190/yr — 2 months free
Register your interest →5 users — add or remove anytime, no seat cliffs
per month · $3,180/yr — approx. $64/user
Register your interest →Founding rate. Available from launch on 31 August 2026. Subscribe before 30 September 2026, or before our second state goes live — whichever comes first — and your rate stays fixed for as long as you're subscribed. More than 15 users, purchase orders, or a government agency? Tailored terms are available. Registered charities, conservation and Landcare groups qualify for discounted pricing.
Split between the Australian Wildlife Conservancy and Wildlife Queensland. BioGeo is a profit-for-purpose business, so the contribution is structural rather than promotional, and continues for as long as the subscription runs.
$5 from every subscription each month, at no additional cost to the subscriber.
BioGeo is a geospatial and ecological science consultancy. SiteCensus was built for work BioGeo does itself.
Contributions are tracked against subscriptions, so the amount is verifiable rather than indicative.
Anything not covered here — email info@biogeo.com.au and we'll get back to you promptly.
SiteCensus queries the same authoritative REST endpoints the agency portals use. The difference is that it runs them together against one study area and produces an assessment document. No government portal spans Commonwealth and State in a single intersect.
Every layer is fetched live from the custodian endpoint at the time of assessment. There is no locally cached data. Update frequency varies by layer and custodian, and is shown per layer in the application.
The library holds 450+ layers. A run queries every enabled layer whose data could reach the study area and skips only those that geographically cannot, so the number is site-specific. Each report states how many were queried and how many returned features, and Appendix A lists every one individually.
By a fixed rule, evaluated top down. The rule is printed in the report with the matching row marked, and the thresholds are not adjustable from the report. Modelled or potential-habitat evidence can inform the narrative but never raises the rating on its own.
It is recorded. A layer that fails its query is reported as failed rather than returned as zero, and a service that cannot be loaded is named in the report with its subject matter marked as not covered.
They are suitable as supporting material in desktop assessments, EIS appendices, flora and fauna assessments and preliminary constraint reports. Results are indicative and require verification with the relevant data custodian before being relied on for statutory decisions.
Yes. The library covers ecological, abiotic, hazard, tenure and planning data. Planners, engineers and government officers use it to understand or independently verify what is on a site without producing the assessment themselves.
No. Outside SiteCensus, BioGeo's work is geospatial science and ecological modelling: corridor connectivity, habitat suitability modelling and remote sensing. BioGeo does not offer desktop ecological assessments, DA support or field ecology.
Screen somewhere your team has done manually, and compare the two.