Cross-check payment application claims against contract terms and site records to help models identify insufficient evidence, missing measurement grounds, and conflicting records.

Extract hazards, exposed entities, and control measures from work-at-height observations to streamline the structuring of safety records.

This dataset is used to identify the material stacking situation at construction sites to improve site management and safety.

Explore planning steps and risk-control evidence for demolition preparation, and evaluate the accuracy and traceability of RAG retrieval.

Connect queries with supporting evidence on key demolition safety topics, including work sequence, structural risks, and stability measures.

A focused collection of heavy machinery and special-purpose vehicle images with category annotations for training visual recognition models and developing engineering applications.

Convert zoning standards into structured JSON that preserves units, applicability conditions, and exceptions for more efficient planning data workflows.

Convert component descriptions and fabrication requirements into structured specifications for prefabricated construction data management and model training.

Evaluate how well models interpret numerical conditions, applicable configurations, and restrictions in scaffolding design and load specification texts.

Verify construction progress claims against reports and planned milestones, with a focus on dates, work status, and conflicting evidence.

Connect crack evolution, structural damage, and repair outcome assessment to help models learn verifiable maintenance decisions for concrete structures.

Supports underground drainage inspection and defect recognition with annotated pipe images and accompanying metadata.

Trace structural response from member failure to progressive collapse and support collapse assessment and intervention strategy training.

Train models to translate land use planning controls with aligned texts and domain-specific terminology.

Train models to verify construction conclusions more reliably by linking submittal content to project requirements and supporting evidence.

The Bridge Surface Damage Image Segmentation Dataset is used to detect and segment various damages on bridge surfaces.

Train models to assess development compliance by connecting parcel conditions with layered zoning rules.

Pair concrete structure imagery with defect annotation paths and category metadata to support visual diagnosis models and inspection workflows.

Pair commercial construction opportunity notices with structured JSON to train models for accurate field extraction and schema-compliant output.

The excavation site robotic arm action control dataset enhances the efficiency of intelligent control in the construction and real estate industry.