By role
AI for landscape architects: planting data, quantities and coordination
A useful landscape workflow connects the visual concept to planting locations, quantities and current schedules. Start with identified objects and an agreed counting rule.
Contents · 6
Turn the concept into identified decisions
AI can help a landscape architect organize a site brief, compare visual directions and reconcile planting data. For production work, each accepted decision needs a location, identifier and source. A plausible garden image does not supply that structure.
When comparing visual options, hold paths, entrances, retained trees and service access fixed. Record which features are illustrative, including season and planting maturity. Carry accepted choices into the planting plan and schedule; leave incidental details of the image outside the commitment to the client.
Translate broad intentions into questions. A “green screen” needs a defined available strip, desired seasonal effect, maintenance regime and relationship to nearby uses. An assistant can structure those questions. Plant selection follows the site's evidence and the landscape specialist's judgment.
Build a site and plant evidence register
Collect location, exposure, shade, wind, soil information, drainage, irrigation provision and maintenance constraints. Mark missing survey or test data. Neither an image nor an enthusiastic species description establishes soil conditions.
For candidate plants, retain botanical name, cultivar, source, mature dimensions, specified supply size and review status. Use project-relevant horticultural references and local experience. A climate rating alone does not describe every site's exposure, drainage or establishment conditions.
Check nursery availability separately from suitability. A catalogue entry is not confirmation of stock, size, quality or delivery date. When a substitution is proposed, compare the characteristics the project needs and identify affected spacing, quantities and maintenance assumptions. Do not silently replace a cultivar with a similar name.
The useful AI output is a structured candidate register with open checks. It should not turn unverified properties into accepted specification fields.
Define what each drawing object represents
One symbol may mean one tree, one shrub group or an existing canopy. Count physical plants using the agreed data, rather than treating all blocks as single items. Keep existing, retained, transplanted, removed and proposed plants in separate selections tied to the survey.
| Item | Counting basis | Evidence |
|---|---|---|
| Individual tree | One approved planting ID | Location and schedule row |
| Shrub group | Explicit group quantity | Group ID and quantity attribute |
| Area planting | Area and approved density | Boundary and counting rule |
| Paving | Defined surface area | Closed boundaries and deductions |
| Furniture | Approved mark and placement | Location, footprint and interfaces |
A table export may be enough for reconciliation. Geometry requires a drawing reader, checked script or configured CAD connector. A PDF remains useful for reviewing plotted labels and context, but it does not automatically expose object identities. See AI and DWG for preparing that route.
Worked example: three counting methods
Suppose a courtyard plan has eight proposed trees, three shrub groups containing 12, 9 and 15 plants, and two beds of 18 and 24 m². The landscape architect has already approved a density of five plants per square metre for this illustrative exercise.
Use only the supplied revision B planting data. Count individual trees by unique planting ID, shrubs by each group's declared quantity, and bed planting by net area multiplied by the supplied density. Return separate totals with source IDs, units and exclusions. Do not select species, change densities or merge existing and proposed planting. List missing or duplicated IDs.
The expected totals are eight trees, 36 shrubs and 210 plants in the area-based category: 42 m² × 5 plants/m². Counting three group symbols as three shrubs would be an error.
Check the included-object map. Confirm that the eight tree IDs exclude existing trees and duplicate symbols on enlarged details. Independently check group attributes and bed boundaries, including deductions.
If one bed shrinks by 4 m², the same rule reduces the area-based quantity by 20. The result must identify the changed bed and schedule row. Check the physical spacing as well: area multiplied by density can be an initial estimate, while a narrow or irregular bed may need explicit placement and a documented quantity adjustment.
Coordinate planting with the rest of the site
Review trees against the current services information, furniture against access and maintenance zones, and beds against grading and drainage. A symbol overlap can prompt a question; resolving it requires the actual dimensions and constraints.
For irrigation, the assistant can reconcile zone IDs and equipment references. That operation does not establish available pressure, hydraulic performance or the suitability of the water source. Transfer current planting boundaries and requirements to the responsible designer.
Keep a coordinated revision record for the survey, planting plan, schedule, paving and irrigation brief. When a plant or boundary changes, identify all affected documents. Close the issue against the reissued package, not just the revised total.
Start with a checkable ARAI task
ARAI can inspect supported DWG/DXF geometry and prepare reviewable outputs. Prepare a landscape drawing task. Review the unsent draft, attach the current file and specify units, object meanings and exclusions. Ask for selection evidence before relying on quantities; plant suitability and system design remain separate reviews.
Ask ARAI about your project
Prepared by ARAI’s editors with AI assistance. Product information was checked against official sources as of the update date. Report an error at hello@arai.build.