Synthetic IFC Benchmark for Rail IFC Quality Gating (R3–R7) (v1.0.0)

This record provides a fully synthetic, non-sensitive IFC benchmark suite to support open benchmarking of deliverable-level quality auditing and publish/hold gating for rail assets. The benchmark is designed to stress-test common failure modes targeted by the R3–R7 metric suite, including: spatial regime inconsistencies (projected UTM/metres vs local vs unit/scale mismatch), TrueNorth outliers, inter-model anchor dispersion, semantic degradation via proxy overuse, and staged information completeness failures in core PSET profiles.

The release includes: (i) a set of synthetic IFC files, (ii) a benchmark manifest describing each case and injected fault scenario, (iii) expected outputs (file-level gate outcomes and failure drivers) to enable automated regression tests, and (iv) benchmark profiles/threshold files and a benchmarking protocol for cross-tool evaluation.

The benchmark is intended for research and method comparison (cross-tool / cross-owner / cross-jurisdiction) and does not represent a real depot or operational asset model.