How to get efficient data management from BIM


Driving efficient data management from BIM and how to become a digital enterprise is examined by Graeme Forbes of Clearbox

As we continue forward on the digital journey how far do we go until we step back and re-examine the outcomes to question the process and the objectives that got us there?

One of the best outcomes I have experienced from stepping out of 30 years delivering design and build projects, into the world of technology, is the new visibility I have gained working with companies and their aspiration of a digital outcome.  Of course, this visibility is all the more informative because it is supported by my exposure to the time I spent faced with the same problems delivering design and build projects around the world.

The core objective of creating the object data in BIM is to provide better information, knowledge and wisdom to allow us to make better decisions, but equally, the opportunity to drive a more efficient process.  As a young engineer in construction back in the eighties, project’s employed staff on sites to undertake time and motion studies. These people determined the man-hours for tasks and looked for savings in time. The studies were done because the contractor paid for the resources and owned the outcome. Get it right and you made money, get it wrong and you lost money. However in today’s contract landscape we tend to simply sub-contract out the work so the issue is not so much the efficiency but the price of the work and whether we can bid and win the project with the prices we receive from the market. That doesn’t mean that projects don’t go wrong. It is more that we are one step further removed from the front line with the result that the skills we need have changed and our sensitivity to the efficiency of the process has changed.

In design people had metrics. Some were based per drawing or based per m2 of the building, and these yardsticks have continued for years.  Along comes BIM and the process changes, and we are now in transition into whatever shape the new world will look like. We are in a state of flux partly because we have the complication of BIM added as a bolt-on while we determine what the longer-term solution to a digitised workspace really looks like. 

The information delivery process 

I am sure many will return to time and motion analysis to review their new processes, but now it will no doubt be under the auspice of lean and six sigma processes.  Of course, these techniques are applied to the existing process to look at any step that does not add value with the intention of removing that step from the process. So, for example, we look at the process of adding data into models, and we realise that to do this in the model environment is inefficient, so we extract the data with a bolt-on tool, edit in Excel and then send the data back into the model. The tool provides the work-around but the process is still manual and unmanaged, and the view of the data limited to what we see when we extract the data. Of course, the process also involves two people — the person with the authoring tool extracting the data in a table, and the person who then adds and edits the data to give back to the operator of the authoring software to re-enter the data.

The information we need to attribute is a mix of data and documents and extends from design right through to the digital handover.  As the process starts to run across the contractual boundaries, we have the complex issue of who edits what in the model files, and we see many contractors employing the designer to make the edits of data. This is either because the authoring tools are too complex to use, or because people are concerned with liability and the confusion of who did what in what model.

Managing the data brings new challenges and opportunities 

So we have issues of efficiency, liability, double handling of information, and ultimately, whether the model can handle the data — let alone the attachment of documents. But the problems don’t stop there. Now that the model is full of data, how do we control what we share with others? While collaboration is great, sharing the cost and which cost; sub-contractors cost, contractors cost, the cost of replacement or the cost of supply or the asset value for the balance sheet, all represent different outcomes from the same core information.

Furthermore, we need to address security. This is the wild card in the BIM journey because while some need to know about the gas pipeline under the road that supplies most of the UK’s requirement, or the aviation fuel network that feeds Europe’s airports, we don’t need everyone to know.  So excluding objects from view, and/or their attached data, or indeed all the data of a particular type, e.g., cost, places considerable pressure on our processes given the abilities of our tools. I am sure a good six sigma process would make a difference, but we have been driven to design our processes based on the existing tools instead of designing the right process and building the tool to suit.

Design the tools to suit the required process– the benefits are much greater

At Clearbox we built our solution to manage a better process, not the process of design, more the process of documenting and detailing the design so that the information was held centrally as the ubiquitous single source of information. This solution allows all parties to access controlled entry to a contextualised view into the information. It would be wrong to suggest that we were so far forward in our vision that we had seen all the problems coming because we didn’t. Good fortune and careful responsive architecture has allowed us to respond to many issues and shown the value in our approach, not least because while it is compatible with more traditional BIM processes, it is more efficient and more effective.

When we started out using a well-known viewer with an API to control the viewer, we didn’t fully realise that the viewer would both run out of power and functionality to perform the simple but sophisticated tasks we were looking to undertake. It was still a viewer we needed, but one that had real scalability, speed, and the ability to run on the PC and access on site. So, we built our own, and now not only do we see it outperform what we had been used to, but we have the future scalability and performance contingencies to cover the next phases of the journey — the fully loaded data and document rich infinite model.

Data federation makes it easier to achieve the desired outcome 

Architecting the solution to deliver data management from the ground up, federating the BIM around the data and using models as containers to deliver the geometric objects to the central database overcomes many other issues. The models remain light, change control operates automatically at an object level with the author of the change identified, and all parties can take part, subject to their access rights.  Unlike BIM, our solution is fundamentally a multi-user environment. In the critical phase of data attribution to models, we can see performance levels improved by a factor of 10 alongside a dramatic reduction in the volume of data moved in each interaction.  So this is now about the benefits of data management via models versus data management via a database solution like BIMXtra. This should all be agnostic of the authoring software used or indeed the combination used on any one project.

Outputs are easier from structured data 

If it were all about the efficiency, that would be one thing, but it is also about data enablement.  We need data attribution for COBie, but we also need data attribution for procurement, for construction, for 4D, for work packaging amongst all other tasks and these attributes don’t necessarily follow the defined attribute structure outlined for COBie.

Bringing consistent structure to the attribution to enable all these other tasks essentially requires this additional attribution to be predefined at the outset for each author and each authoring tool if the task is to become scalable — not to mention be used across multiple projects to provide consistent information.  By ensuring compliance of the information entering at the beginning from many different sources and by allowing that information to automatically map to a defined structure, the only requirement a tool like BIMXtra places on the design consultant is to work consistently. You can work differently in each office (if you must), but you should work consistently — a requirement ultimately required of nearly any other tool. Once the data is federated, enhanced and authorised for use in line with the BS1192 process, outputs can then be configured for output.  COBie is pre-set as an output, but other formats are configurable and can be saved to be re-used. Of course, these outputs extract data from the database irrespective of the format of the input, and this now includes configurable IFC outputs amongst the range of other formats, allowing users to take the information and perform the task they need to undertake.

The opportunity is scalable 

BIM provides a massive opportunity to support new and better ways of working. It also benefits from re-examining the outcomes we wish to achieve and the processes and tools we need to deliver a lean process that can be repeated time and time again.  BIM Level 2 in its generic form is one solution, but it is neither the most efficient nor the most effective.  If you would like to join a growing range of customers not prepared to accept the status quo we would be pleased to hear from you and ultimately support you in becoming a digital enterprise.

Graeme Forbes

Managing Director


Please note: this is a commercial profile


Please enter your comment!
Please enter your name here