GIS digitisation & conversion
Paper sheets, scans and legacy CAD brought into a clean, attributed schema your GIS can actually query.
- Input
- Paper / scan / legacy CAD
- Output
- .SHP / .GEOJSON / geodatabase
- Positional accuracy
- Matched to source control
- Schema
- Client-defined attribute fields
Paper sheets, scans and legacy CAD brought into a clean, attributed schema.
Vector layers (.SHP/.GEOJSON)
Features digitised and topologically cleaned — no dangling nodes or overlapping polygons left for your team to fix.
Attributed schema
Fields defined against your existing data dictionary, not a generic default schema you have to remap.
Georeferenced source scans
Original sheets rectified and delivered alongside the vectors, for provenance and QA.
No surprises between quote and handover.
- 01
Source assessment & georeferencing
Scans, sheets or legacy CAD rectified to a known coordinate system before a single feature is drawn.
Rectified source · control report - 02
Schema definition
Attribute fields agreed against your existing data dictionary so the output slots into your GIS without remapping.
Schema spec - 03
Digitisation
Features traced and attributed, with topology rules enforced as they're drawn rather than fixed after the fact.
Draft vector layers - 04
Topology QC & handover
Layers checked for gaps, overlaps and dangling nodes before delivery.
Clean SHP/GEOJSON · QC log
Who this is for
- Urban planning
- Power & utilities
- Infrastructure & rail
- Water resources
Questions worth answering upfront.
- Can you work from paper maps that have no coordinate information at all?
- Yes — sheets are georeferenced against known control points (road intersections, survey marks, or a coordinate grid printed on the sheet itself) before digitisation starts.
- What file formats can you deliver into?
- Shapefile, GeoJSON, or a file/enterprise geodatabase — whatever your existing GIS ingests, matched to the attribute schema you already use.
- How do you handle inconsistent legacy CAD layers?
- Legacy layers are reviewed against your target schema first; features get reclassified and re-attributed during digitisation rather than passed through with their original, inconsistent layer names.

