The philosophy is that one should weed out the errors and mistakes commonly found in construction projects in the virtual world where it is cheap, instead of at the construction site where it is very expensive.
If you can understand Danish you can learn more here: MT Højgaards lille revolution
I would agree that this approach will optimize and help a lot in the actual project. It is in my view a big step in the right direction. However, big projects today are still executed in “silos”. There is very little communication between projects, re-use of concepts, knowledge, data or documentation. In other words, each project is its own world and one gets very dependent on knowledgeable key people to pull off successful projects. Historically this is not so strange (think about the Master Builders), and the same thing is true for plant related industries, however some interesting changes are happening there.
More and more EPC’s (Engineering Procurement and Construction companies) have moved in a direction where concepts, processes, knowledge, engineering data, documentation, supply and logistics data is shared across projects as a form of structured master data. This is achieved by defining modular and smaller building blocks that are needed in projects, and allowing them to be adjusted in the project specific context. Each project is of course still a unique “one-off” but with more and more re-use, traceability and controlled changes.
So how could this translate to construction industry?
Well, I think a first step would be to manage the portfolio of projects, and BIM models together in some sort of information backbone. This could be a PLM like solution (There are examples of those) where data can be analyzed, managed and executed with process support across all projects. You can read more about this in one of my earlier blog posts: “VDC is that like PLM for construction industry”. Such a solution would also brings control to the thousands of documents that are still specifying to the projects and the BIM models themselves. An approach like this would take a big step towards “linked or structured data”
Another approach could be more federated, however it would still require some sort of engine to orchestrate data and information from all systems involved in all projects.
If you’d like to read more about a data oriented approach, I would recommend these 3 articles written by Jos Voskuil
“The difference between files and data-oriented – a tutorial part 1”
“The difference between files and data-oriented – a tutorial part 2”
“The difference between files and data-oriented – a tutorial part 3”
The image used in this post is by Adam121 and purchased at dreamstime.com