SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsaec-datasets-tools

aec-datasets-tools

Writing & research

View repository

Updated 10-15 09:49

About this skill

AEC Research Datasets & Tools — Reference

Point researchers to credible tools and (where they exist) open datasets/benchmarks, and remind them to cite software and data (version + DOI) and to verify licenses and current availability before use.

1. Building energy & performance

  • EnergyPlus (DOE) — whole-building energy simulation (the research standard); OpenStudio SDK/GUI; EnergyPlus + Python (eppy, pyenergyplus), besos, eppy for parametric runs.
  • Modelica ecosystem — Buildings library (LBNL), Spawn-of-EnergyPlus, Dymola/OpenModelica.
  • Ladybug Tools (Ladybug/Honeybee/Butterfly) in Grasshopper — daylight (Radiance), energy, CFD.
  • Radiance — daylight/glare; CONTAM — airflow/IAQ; OpenFOAM — CFD (ventilation, urban microclimate).
  • TRNSYS, IDA ICE, IES-VE, DesignBuilder — common (mostly commercial) BES tools.
  • Weather data: EnergyPlus/EPW files (climate.onebuilding.org), TMY3; UBEM: UMI, CityEnergyAnalyst (CEA).
  • Open datasets/benchmarks: ASHRAE Great Energy Predictor III (Kaggle), Building Data Genome Project 1/2, the DOE Commercial/Residential Prototype Building Models, and NREL ResStock/ComStock (verify current terms).

2. BIM / openBIM / interoperability

  • IfcOpenShell (Python/C++) — parse, query, geometry from IFC; BlenderBIM/Bonsai — open BIM authoring.
  • xBIM Toolkit (.NET), IFC.js / web-ifc / That Open Engine — IFC in the browser.
  • Speckle — open data hub / interoperability across AEC tools; Solibri — model checking (commercial).
  • buildingSMART resources: IFC schema, IDS samples, bSDD; Linked Building Data: IFCtoLBD, BOT ontology.
  • Datasets: IFC sample files (buildingSMART, KIT IFC examples). Curated open BIM datasets are scarce — note this limitation; many studies build their own.

3. Computer vision, point clouds & ML in construction

  • Point clouds: PDAL, Open3D, CloudCompare, PCL; photogrammetry/SLAM: COLMAP, Meshroom.
  • Benchmarks/datasets: S3DIS, ScanNet, Semantic3D, KITTI (general 3D/segmentation); construction- specific: SODA / MOCS / ACID (site object/worker detection), Structured3D, building-defect crack datasets (e.g. SDNET2018, METU concrete crack) — verify licenses & current links, and report dataset size/splits honestly (AEC ML datasets are often small/narrow).
  • ML frameworks: PyTorch, TensorFlow/Keras, scikit-learn, Ultralytics YOLO; GNN: PyTorch Geometric/DGL.

4. Structural / FEM / SHM

  • OpenSees (earthquake/structural simulation, open), Code_Aster, CalculiX, FEniCS/FreeFEM (open FEM); Abaqus/ANSYS (commercial). SHM datasets: Los Alamos (LANL) benchmark, Z24 bridge, community SHM repositories (verify).

5. Computational & generative design

  • Grasshopper/Rhino (+ Galapagos, Wallacei, Opossum, Octopus for optimization), Dynamo/Revit, Rhino.Compute, COMPAS (open Python framework for AEC research), OpenSCAD/CadQuery (code CAD).

6. Construction management / simulation / decision

  • DES: SimPy (Python), AnyLogic (also ABM), Arena; ABM: NetLogo, Mesa (Python), Repast.
  • Scheduling/optimization: Python pulp/OR-Tools, pymoo (multi-objective), DEAP, Platypus.
  • MCDM: R MCDA/topsis, Python pymcdm, pyDecision; AHP spreadsheets/ahpy.

7. Bibliometrics, statistics & qualitative analysis

  • Bibliometrics/science mapping: VOSviewer, CiteSpace, Bibliometrix/biblioshiny (R), OpenAlex/pyalex, litstudy, Gephi (see bibliometric-analysis).
  • Statistics: R (tidyverse, lavaan/semTools for SEM, psych), Python (pandas, statsmodels, pingouin), SmartPLS / seminr (R) / plspm for PLS-SEM, JASP/jamovi (free GUI), SPSS/Stata (commercial).
  • Qualitative: NVivo, ATLAS.ti, MAXQDA, Dedoose, Taguette (open), QualCoder (open).
  • Reference managers: Zotero (open), Mendeley, EndNote (see citation-formats).
  • Reproducibility: Jupyter, Quarto/R Markdown, renv/conda/venv, Git, Docker, Binder.

8. Citing software & data (do this)

  • Cite the version and a DOI where available (Zenodo archives releases; many tools have a "how to cite" / CITATION.cff). Treat as @misc/software in BibTeX (see citation-formats).
  • For datasets: cite author/provider, title, version, year, repository, DOI, and access date; state the license and any usage restrictions.

9. Cautions

  • Verify availability & licenses — links and terms change; some "open" datasets restrict commercial/ redistribution use.
  • Report dataset provenance, size, and splits — small, single-source AEC datasets undermine generalizability; say so.
  • Pin versions (software + libraries + weather files) for reproducibility; differences change results.
  • Prefer open, scriptable tools for reproducible pipelines; if using commercial tools, document settings thoroughly so others can replicate the workflow.