by John Millar | March 5, 2021 | 5 min read
In recent years, there has been much talk of Digital Twin and the many potential benefits that it offers to those who own, manage and maintain built assets. The main selling point is often in the valuable data-driven insights it provides, which can inform best practice and support O&M decision-making over the course of the asset’s lifecycle.
Whilst the normalisation of knowledge-driven O&M is certainly an important milestone to work towards as we continue to drive BIM diffusion both in the UK and abroad, the concept carries a number of pragmatic concerns. How are we planning to capture, represent, classify and store (in a structured fashion) the significant volume of miscellaneous knowledge that will aggregate over the several decades of a built asset’s lifecycle? Is it possible to model this knowledge in an environment similar to the one in which we model our information? Could we address this issue with the tools and guidance currently available? If so, how?
One potential solution lies with Building Knowledge Modelling (BKM). A theoretical solution proposed by Motawa and Almarshad (2013), which involves the integration of Knowledge Management (KM) principles into BIM practices, resulting in a technological environment which significantly increases the management power that BIM is able to provide by overcoming one of its main weaknesses; the representation of soft data (qualitative and subjective observations resulting from the user’s cognitive processing of information).
This enhancement allows agents involved in the management of a built asset to capture and retrieve all manner of explicit knowledge relating to the asset and its day-to-day operation. Depending on the specific implementation, it offers several intelligent functions; it could automate the detection of maintenance backlogs and the retrieval of information relating to relevant building elements; inform the prioritisation of maintenance works and identify maintenance relationships and patterns, also allowing the prediction (and therefore, prevention) of common-cause failures.
Taking a long-term view, it allows organisations to align their O&M activities with organisational performance, thus assisting in the achievement of their organisational objectives over time (Wanigarathna et al., 2019).
The BKM concept described by Motawa and Almarshad (2013) comprises of an Asset Information Model (AIM) for the retrieval of building information, a knowledge base and a web-based interface module for the bi-directional exchange of soft data. Within the AIM, a pre-defined set of parameters are assigned to managed and maintainable building elements for the storage of knowledge case details, which the knowledge-based module can identify via an Industry Foundation Classes (IFC) protocol and present for review via an integrated browser function.
Fig. 1 – Knowledge case parameters embedded within a Building Information Model (Motawa and Almarshad, 2013; Motawa, 2015).
By leveraging the power of Case Based Reasoning (CBR, a paradigm of artificial intelligence dealing with problem solving based on past cases), the system will also be able to recognise other elements which are potentially affected by a particular maintenance operation and present similar cases to the one in question via association by element clustering as an intelligent representation of maintenance-related element relationships.
The end result can be thought of as a kind of BIM-driven expert system, with the building knowledge model able to present evidence-based recommendations on how O&M activities can be best conducted, on demand, with its richness of knowledge increasing incrementally over the course of the asset’s lifecycle.
The BKM concept may hold significant potential in light of our industry’s current aspirations towards digitally enhanced operations and a strengthened partnership between those who deliver the built environment and those who manage it. Where a project-specific BKM solution is implemented, the owner’s AIM forms an integral part of a greater data-centric socio-technical system, one which aids the organisation in the achievement of its objectives via pro-active and probabilistic O&M strategy.
On the supplier side, it holds great potential for the enhancement of Post Occupancy Evaluations (POEs), with an efficient and effective means of intelligent, asset-specific knowledge exchange enabling deeper (and more frequent) insights across all projects undertaken, resulting in greatly improved service delivery and driving continuous improvement at an industrial level.
Fig. 2 – System output; captured knowledge retrieved from a previous case (Motawa and Almarshad, 2013; Motawa, 2015).
However, in order to fully benefit from the vast quantities of potential knowledge that we have thus far left uncaptured and hence unutilised, in addition to the ocean of big data we are now preparing to dive into, a fundamental shift in attitude is needed. Maintenance activities should be regarded less as an unavoidable financial burden and more as an opportunity to add value at an organisational level.
As a result, BIM methodology should not be considered as merely an enhancement of the capital delivery process (with some data delivered on the side); it should be recognised as the first step in developing what could ultimately become one of the organisation’s most valuable assets: a digitalised single source of truth for all explicit knowledge relating to the design, construction and operation of their built assets – the vehicles through which they deliver their services and achieve their business aims.
For more information, please contact John Millar.
Member of the Ryder Alliance
+44 (0) 191 269 5444 [email protected]
Subscribe to our newsletter for our latest insights into all things digital.