Big data in design and construction is coming, and it is being fuelled by Building Information Modelling (BIM). While many architecture, engineering and construction firms have been using “BIM” (or parametric modelling, as it was called back then) for years to generate their drawings, the industry is now focused on maximizing the capture and reuse of the “I” in BIM: information. This trend is mainly fuelled by owners who want to realize the benefits of an information rich BIM model to help design, construct and operate their projects.
Traditionally, the AEC industry has thought of big data as that which is generated by construction correspondence on site, which is usually mandated to be in writing. But that is not big data, it is a big records management task. That is why most medium to large projects now make use of an electronic document management system (EDMS). These systems usually only capture and track project records which have been issued (or published) by one stakeholder to other stakeholders. Information however, is a representation of data in a manner which is suitable for making informed decisions. Only through informed decisions can we build and manage high performing assets.
The industry has witnessed an ever increasing use and re-use of digital data in building information models. This is done either directly, for example by means of coordination using BIM or indirectly by exporting BIM geometry and data to, or integrating with, design analysis programs and construction costing and scheduling tools. BIM is also extensively used for scenario analysis and evaluation of design and construction options in a way similar to rapid prototyping in the manufacturing industry. Furthermore, BIM provides a repository for facility management information and is being specified by owners and operators as a means by which operational asset data should be delivered at project handover.
Although the size of data generated in a BIM process during a project’s design and construction phase might not equal a fraction of that generated by websites and applications like Facebook in a single day, the size of the data created and captured in a BIM process will only grow. This is particularly true as BIM processes expand to include facility and asset management. But despite the relatively small volume of data, BIM data does exhibit other characteristics of big data, namely variety, velocity and variability (or veracity).
A BIM process incorporates a variety of data, from initial concept massing to material properties and object costs and codes. Once models are created more data, such as drawings, are produced quickly. Further, with parametric design iterations, thousands of data points and design options can be created with velocity exceeding that of websites and applications.
There is also variability as different file format are used and data created for different purposes, by different teams. Volume, Variety, Velocity and Variability are often defined as the characteristics of big data. Yet big data is useless unless we know how to derive value from it.
As the construction industry adapts to the paradigm change of a data-driven culture, we must change the way we think of and value the data we produce and capture about our built assets, both during the capital delivery phase and the operations phase of the asset’s life-cycle. We need to apply a framework for collecting, displaying, checking and disseminating data, a consistent management processes which also asks for the information requirements, as early on as possible, of all stakeholders throughout the life of the asset, and not just the capital delivery phase.
In the UK, the framework which provides the foundation of the BIM process is BS1192:2007 and more specifically the Common Data Environment (CDE). Pioneered through the early 90’s on a number of BAA projects, the CDE is a process that ensures all data and information is checked, approved and authorised before it is used by other stakeholders. It enables the extraction of information as views of multi-authored data and through the re-use of validated data it eliminates wasteful activities such as duplication of effort and re-work.
The recently published PAS1192-2:2013 and PAS1192-3:2014 build on and complement the BS1192 processes for collaboration and data re-use; both providing frameworks around the use of the CDE for collecting and disseminating project and asset data. PAS1192-2 focuses on the capital delivery phase, defining the information management requirements, maximising the re-use and spatial coordination of the project data and information to create a valuable, digital representation of the assets which can, at hand-over, be used by the Employer (asset owner) and operator to efficiently operate and maintain their assets without the need to recapture data and information at a later date, and at extra cost.
On the other hand, BS1192-3 focuses on the operational phase of an asset; it provides a framework for information management for the whole life-cycle of asset management. This framework supports the operational stage of assets irrespective of whether these were commissioned through major works, acquired through transfer of ownership or already existed in an asset portfolio. PAS1192-3 was developed in recognition of the fact that the cost of operating and maintaining buildings and facilities can represent up to 85% of the whole-life cycle cost of an asset.
Ownership of the CDE and the requirement for specifying the use of the CDE processes should sit firmly with the employer. After all it is the employer who owns the asset, the data about it and who will take receipt of the data captured throughout the capital delivery phase, including verified and validated as-built data, for reconciliation with, or creation of, their Asset Information Model (AIM). If at any point the asset owner changes then so does the ownership of the CDE.
With ownership of the CDE the employer can ensure that processes to manage, maintain and operate the asset are implemented. Furthermore, the ‘occupier’ or ‘end user’ (if not the employer) should have suitable access to the CDE, enabling them to contribute to the data through general maintenance activities, providing the employer with an accurate and current Asset Information Model throughout the asset’s life.
Unfortunately many employers, particularly in the Middle East, satisfy themselves with a visual briefing of a coordinated 3D model during the capital delivery phase, and an as-built 3D model (albeit with “information”) at project handover. However, only through proper understanding, capture and management of data, most of which will be generated by the BIM process, can we realize the true benefits of Building Information Modelling and manage the big data that will be generated on our projects.