Precision Land Partitioning Engine
Deterministic multi-objective parcel subdivision formulated on computational geometry and the Pareto frontier.
The 6-Stage Partitioning Pipeline
01 // Boundary Topology Ingestion
Input parcel vertices are validated for planar closure, orientation, and self-intersection. High-magnitude coordinates are projected into a LocalMetricFrame.
02 // Constraint Specification
Definition of legal and topographic constraints: target lot counts, minimum area thresholds, road access frontage, and alignment vectors.
03 // Candidate Generation
Iterative generation of candidate subdivision lines and road networks using deterministic geometric decomposition.
04 // Multi-Objective Pareto Evaluation
Scoring of candidate configurations across competing objectives: road length minimization, isoperimetric compactness, and area equity.
05 // Topological Verification Gate
Strict non-overlapping validation, boundary conformance checks, and machine-precision area residual tracking.
06 // Cadastral & CAD Export
Emission of validated parcel boundaries to DXF, Pregeo DAT cadastral libretto format, and openBIM IFC site geometries.
Multi-Objective Formulation
Mathematical balancing between competing design parameters on the Pareto frontier.
Minimization of internal road network length to reduce infrastructural costs.
Maximization of isoperimetric compactness (4*PI*Area / Perimeter^2) to prevent unusable elongated lot geometries.
Minimization of area deviation variance relative to target entitlement shares.
Enforcement of minimum legal frontage requirements along existing or planned road access.
Natural Language -> Constraint DSL AST
Natural language expressions are translated into a formal Abstract Syntax Tree without substituting the deterministic core.
// QismaCAD Formal Constraint DSL AST
BOUNDARY poly_502 {
COORDINATES: LOCAL_METRIC_FRAME;
SURFACE_AREA: 5748.25m2;
}
CONSTRAINTS {
TARGET_UNITS: 4;
PROPORTIONAL_SHARES: [0.25, 0.25, 0.25, 0.25];
MIN_FRONTAGE: 12.0m;
ROAD_RESERVE_WIDTH: 5.0m;
MANDATORY_STREET_ACCESS: true;
}
SOLVER_WEIGHTS {
ROAD_MINIMIZATION: 0.40;
ISOPERIMETRIC_COMPACTNESS: 0.35;
AREA_EQUITY: 0.25;
} The natural language input is parsed into a structured AST with strongly typed parameters. The mathematical solver evaluates strictly the formal constraint DSL, guaranteeing reproducible results.
Natural language text parsing operates strictly at the UI layer. The underlying computational geometry core evaluates the typed AST without floating-point heuristic ambiguity.
Geometric Boundaries & Infeasible Sets
Transparent declaration of geometric validity limits and computational constraints.
Not every geometric boundary admits a feasible partitioning solution. Severely concave polygons, acute internal angles (<15 deg), or mutually exclusive constraints (e.g. frontage requirements exceeding boundary perimeter) are rejected with formal diagnostic diagnostics rather than unconstrained approximations.