NSF EAGER/Collaborative Research:Science-Based Exploration of Invariant Signatures of Architecture/Engineering/Construction Objects to Enable Interoperability of Building Info Modeling

The invariant signature of an architecture, engineering, and construction (AEC) object is defined as "a set of intrinsic properties of the object that distinguish it from others and that do not change with data schema, software implementation, modeling decisions, and/or language/cultural contexts." (Wu et al. 2021). To explore invariant signatures, we collected industry foundation classes (IFC)-based building information models (BIMs) and extracted AEC objects from these models. We then labeled each IFC object with its corresponding category of beams, columns, footings, slabs, and walls. This dataset can be found in Purdue University Research Repository (Zhang and Wu 2019a). We've developed this dataset containing 1,891 AEC objects, including 795 beams, 412 columns, 348 footings, 74 slabs, and 262 walls, collected from 5 IFC instance models. A wide range of common shapes were covered. For example, for beams, the dataset covers C-beam, I-beam, L-beam, U-beam, beam with cuts, and Skewed I-beam (Wu and Zhang 2018). This dataset contains geometric and locational signatures of each object, in addition, we also provided metadata signatures, which are properties associated with the data type standard (e.g., IFC properties). To help analyze geometric representations of AEC object, an interactive visualization of the formation of fundamental 3D representations of a selected architecture, engineering, and construction (AEC) object was created in game simulation in a first-person view (Zhang et al. 2019). An animation showing an example of such visualization of a cone frustum-shaped bridge pier can be found below:

In addition, an empirical data-driven approach was proposed for "achieving a systematic understanding of entity definitions in an IFC schema. The approach utilizes IFC data and schema in a synergistic way, to facilitate such systematic understanding. Experimental testing is used to serve as verifications of the understanding and accrue the understanding, along with which byproduct BIM tools will be developed. The proposed approach was tested on understanding entities for geometric representations in the IFC2X3_TC1 schema. Through the experimental testing, systematic understanding of 62 IFC entities were obtained, and a visualization algorithm was developed and implemented based on this understanding." (Zhang 2018). This approach is likely to open the door to IFC tool development in academia. 

The invariant signatures were built upon Cartesian points-based geometric, relative location and orientation, and material mechanical properties, as two main components: invariant geometric signature and invariant material signature. Geometric information of invariant signatures were developed to include the following properties: "(1) number of subcomponents, (2) number of faces, (3) cross-sectional profile, (4) extrusion direction, (5) dimensional ratios, (6) number of straight lines and curves, (7) boundary line connection angles, (8) lengths, and (9) turn directions." "The locational information is reflected by the x, y, z-coordinates of the origin when placing an object as well as its orientation. The material information includes a set of six parameters: material strengths, mass density, Poisson ratio, shear modulus, thermal expansion coefficient, and Young’s modulus for structural analysis." (Wu et al. 2021). An example geometric signature for rectangular parallelepiped shape can be found in (Wu and Zhang 2019a). Example material properties of material signatures can be found in (Wu et al. 2021). Classifiers were built based on the invariant signatures that could automatically classify an input IFC object into beams, columns, footings, slabs, and walls, and classify a beam into rectangular beam, C-beam, I-beam, L-beam, U-beam, round beam, rectangular beam with cuts, truss, hollow round beam, and skewed I-beam (Wu and Zhang 2019b).

The invariant signatures were tested in automated quantity takeoff scenario and automated structural analysis input generation scenario. A Data-driven Reverse Engineering Algorithm Development (D-READ) method was developed for developing interoperable quantity takeoff (QTO) algorithms using IFC-based BIM (Akanbi et al. 2020). This D-READ method "enables the development of QTO algorithms for IFC-based BIMs resulting from different BIM authoring tools and workflows, and therefore it enhances robustness of BIM-based QTO. It takes a novel bottom-up approach in QTO algorithm development compared to the traditional top-down approach. A model view definition (MVD) model for IFC model checking was developed and incorporated with the QTO algorithms. The proposed method was tested on nine different BIM instance models from different sources. A comparison with the state-of-the-art commercial software showed consistent QTO results, whereas the proposed D-READ method resulted in QTO algorithms that were more robust with regard to the different BIM authoring tools and workflows used." (Akanbi et al. 2020). This method enables small businesses and freelancer programmers to practically develop automated QTO algorithms, which lowers the entrance barriers in this research domain and market segment. This can provide a more diverse set of tools for AEC industry to pick from, and lead to better and more affordable access of BIM tools especially for small businesses owners in the AEC industry. "The invariant signatures and the data-driven method were tested in developing the interoperable BIM support tool for structural analysis through an experiment. Ten models were created/adopted and used in this experiment, including five models for training and five models for testing. An information validation and mapping algorithm was developed based on invariant signatures and training models, which was then evaluated in the testing models. Compared with a manually created gold standard, results showed that the desired structural analysis software inputs were successfully generated using the algorithm with high accuracy." (Wu et al. 2021).

The results from this project are likely to make an impact on society by unleashing the full potential of building information modeling (BIM) by enabling a seamless and universal interoperability of BIM. Although the adoption of BIM in the AEC industry rapidly increased in recent years, the lack of interoperability between different BIM software intended for different tasks and phases of an AEC project stood in the way of achieving the highest potential of BIM to support life cycle information needs to all parties and all phases of an AEC project. The developed invariant signatures in this project fundamentally change the way BIM interoperability is addressed, which breaks interoperability barriers caused by data schema, software implementation, and/or language/culture contexts, and makes a smooth BIM-based workflow throughout the life cycle of an AEC project achievable. Such a smooth workflow is expected to help the AEC industry save time and cost, improve quality and productivity.

 

Acknowledgments:

The authors would like to thank the National Science Foundation (NSF). This material is based on work supported by the NSF under Grant Nos. 1745374 and 1745378. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.

 

References:

Zhang, J., and Yang, F. (2024). "Building a Bridge between Building Information Modeling and Digital Twins: Introducing Invariant Signatures of Architecture, Engineering, and Construction Objects." Chapter 2, Digital Twins in Construction and the Built Environment, ASCE, Reston, VA. 

Akanbi, T., and Zhang, J. (2023). "IFC-Based Algorithms for Automated Quantity Takeoff from Architectural Model: Case Study on Residential Development Project." Journal of Architectural Engineering, 29(4), 05023007.

Wu, J., and Zhang, J. (2022). "Model Validation Using Invariant Signatures and Logic-Based Inference for Automated Building Code Compliance Checking.” Journal of Computing in Civil Engineering, 36(3), 04022002.

Wu, J., Akanbi, T., and Zhang, J. (2022). “Constructing invariant signatures for AEC objects to support BIM-based analysis automation through object classification.” J. Comput. in Civ. Eng., 36(4), 04022008.

Li, H., and Zhang, J. (2022). "Improving IFC-Based Interoperability between BIM and BEM Using Invariant Signatures of HVAC Objects." Journal of Computing in Civil Engineering, 37(2), 04022059.

Li, H., Zhang, J., Chang, S., and Sparkling, A. (2022). "BIM-based Object Mapping Using Invariant Signatures of AEC Objects." Journal of Automation in Construction, 145(January 2023), 104616.

Wu, J., Sadraddin, H.L., Ren, R., Zhang, J., and Shao, X. (2021). “Invariant signatures of architecture, engineering, and construction objects to support BIM interoperability between architectural design and structural analysis.” J. Constr. Eng. Manage., 147(1), 04020148.

Wu, J., Zhang, J. (2021). Invariant Signatures of Architecture, Engineering, and Construction (AEC) Objects in Industry Foundation Classes (IFC)-based Building Information Modeling. Purdue University Research Repository. doi:10.4231/7VW7-0129

Ren, R., and Zhang, J. (2020). “Comparison of BIM interoperability applications at different structural analysis stages.” Proc., ASCE Construction Research Congress, ASCE, Reston, VA, 537-545.

Akanbi, T., Zhang, J., and Lee, Y.C. (2020). “Data-driven reverse engineering algorithm development (D-READ) method for developing interoperable quantity takeoff algorithms using IFC-based BIM.” J. Comput. in Civ. Eng., 34(5), 04020036.

Akanbi, T., and Zhang, J. (2020). “Automated design information extraction from construction specifications to support wood construction cost estimation.” Proc., ASCE Construction Research Congress, ASCE, Reston, VA, 658-666. 

Ren, R., and Zhang, J. (2021). “A new framework to address BIM interoperability in the AEC domain from technical and process dimensions.” J. Adv. Civ. Eng., accepted.

Ren, R., and Zhang, J. (2019). “Model information checking to support interoperable BIM usage in structural analysis.” Proc., 2019 ASCE International Conference on Computing in Civil Engineering, ASCE, Reston, VA, 361-368.

Zhang, J., and Wu, J. (2019). Building Information Modelling (BIM) data repository with labels. Purdue University Research Repository. doi:10.4231/60V2-PJ72

Wu, J., and Zhang, J. (2019a). “Introducing geometric signatures of architecture, engineering, and construction objects and a new BIM dataset.” Proc., 2019 ASCE International Conference on Computing in Civil Engineering, ASCE, Reston, VA, 264-271.

Wu, J., and Zhang, J. (2019b). “New automated BIM object classification method to support BIM interoperability.” J. Comput. Civ. Eng., 33(5), 04019033. 

Zhang, J., Chen, Y., Liu, R., and Debs, L. (2019). “Cartesian points visualization in game simulation for analyzing geometric representations of AEC objects in IFC.” Proc., 36th Intl. Symposium on Automation and Robotics in Construction (ISARC 2019), I.A.A.R.C., iaarc.org.

Akanbi, T., Zhang, J., and Lee, Y.C. (2019). “Automated item matching and pricing (IMP) for wood building elements to support BIM-based wood construction cost estimation.” Proc., 2019 ASCE International Conference on Computing in Civil Engineering, ASCE, Reston, VA, 402-409.

Ren, R., Zhang, J., and Dib, H.N. (2018). "BIM interoperability for structural analysis." Proc., ASCE Construction Research Congress, ASCE, Reston, VA, 470-479. 

Wu, J., and Zhang, J. (2018). "Automated BIM object classification to support BIM interoperability." Proc., ASCE Construction Research Congress, ASCE, Reston, VA, 706-715.

Zhang, J. (2018). "Towards systematic understanding of geometric representations in BIM standard: an empirical data-driven approach." Proc., ASCE Construction Research Congress, ASCE, Reston, VA, 96-105.