AEC software is built for buildings; landscape inherits the leftovers. This is where I build the parts that were left out — and carry the design judgment all the way down into the code.
A site arrives as noise — levels, trees, desire lines, code. The work is to read it as relationships, then run one spine of logic through the lot until it reads as a single idea. The method travels; the projects stay with their owners.
↗BNSW school projects live or die by the EFSG standard. I turned its 226 cross-referenced clauses into a working checklist — vague guideline in, executable design decision out.
Stormwater gets treated as a nuisance to drain away. Read it instead as a resource to design with, and it starts organising the site for you.
↗DThe EFSG asks for 35% canopy. Add up crowns naively and the overlaps double-count — so I take the boolean union instead. The abstract target becomes a number you can defend.
↗E40/60 soft to hard, 15% deep soil — easy numbers that fight each other on a real site. The method is making them all hold at once, without the design going rigid.
Revit was drawn for buildings. RevSphere bends it back toward landscape — plan overlays, material legends, floor work and canopy coverage, with more on the way.
↗2The paperwork no design app will touch — batch renaming, mail date-stamping, email-thread archiving — in one calm desktop shell that never deletes a thing.
↗3Project management that actually speaks AEC — landscape stages, blocked-vs-on-hold, DA and IFC milestones — on a board built for one studio's real flow.
Every build starts at one concrete pain point — a workflow nobody automated, a clash buried in the standards. Name it honestly before drawing anything.
Draw the line between two languages — the judgment of landscape, translated into the logic of code. This is the part almost nobody can do from both sides.
It has to land as something real — an app or plugin in actual use, not a slide. Thinking that can't be built doesn't count.