Excitech consultant Rob Clark was part of a committee that recently ran a workshop called BIM4Real to share and solve some of the problems real Building Information Modelling projects encounter. In this article he shares the findings and some personal thoughts.
BIM4Real eschewed the tried and tested conference format by focussing on a fictional building project: the construction of a large shop in the English and Welsh border town of Shrewsbury. The completely fictional client had produced an Employer’s Information Requirements document in line with BSi PAS1192:2. Attendees were split into five groups and included people from different disciplines. Their task was to discuss how they would deliver this project as a team.
There were three focused sessions: project set up; design delivery; and using the data produced to aid the construction and subsequent operation of the building.
Five chair-people summarised at the end of each session what had been discussed. Each table was laden with documentation including industry protocol documents and templates for execution plans as well as the Employer’s Information Requirements document. In each session, the chair-people were asked to get answers to very specific issues that face our industry in the delivery of projects using co-ordinated Building Information Models (BIMs).
What came out proved very interesting. There was plenty of consensus, a lot of confusion and quite a few challenges.
Project kick off
Despite years of under-utilisation the BS1192:2007 Code of Practice workflows were being adopted by more than half of the room. This is likely to be due to the natural standardisation of information that comes from adopting a Building Information Modelling process.
Prior to the meeting I had expected a lot of conversation on liability, but interestingly there was a good amount of consensus that project indemnity covers these requirements. Where there was a lot of debate, however, was model ownership. Initially it was suggested simply that the ‘client owns the model’. This raised a lot of discussion over where the limitations of ownership occurred.
There was concern that simply communicating such ownership would imply that it can be used without restriction, including on other similar projects. Model ownership, it was felt, is for the particular project asset alone.
Apparently, a client saying ‘we want BIM, can you deliver?’ in the qualification is not really sufficient. I have to raise a smile at that rather obvious piece of common sense; however it is plain that a good number of jobs start with this very requirement alone.
There was general acceptance that the new PAS1192:2 document offers good guidance for project set up, procedure and delivery for both the Project Information Model (PIM) and the Asset Information Model (AIM). The single Employer’s Information Requirements (EIR) document was seen as essential to the successful delivery of a project.
A Digital Plan of Works was deemed necessary, including documented and agreed deliverables. This should include all necessary asset information, or indeed if any is needed. The asset team should be involved at the project commencement.
Participants felt it was very important that they understood why information requirements are needed; what is it going to be used for? In order to establish a process to deliver it in the format required.
Is it sensible that asset registers like Construction Operations Building Information Exchange (COBie) are stipulated in industry level process? Wouldn’t it be better to tell the team in pre-qualification what Computer Aided Facilities Management (CAFM) system was in use and what information was required?
The team could then make a format directly available for this.
Is a geometric model required at the end of the process for handover? If so what format should it be in, and again, what will it be used for? Do we understand the expectation here? We should be very wary of agreeing to ‘as-built’ unless we understand what is involved to produce it.
In response to these requirements it was generally accepted that the delivery team needs to assemble a BIM Execution Plan (BEP). There was some debate over the complexity of how this is presented in the PAS1192:2 document, with many suggesting that a single document was required to show execution and delivery.
It was felt that several documents would lead to unnecessary paperwork and confusion. As much as there is merit in two execution plans for pre-qualification and post-qualification and a separate Master Information Delivery with separate task team assessments and IT assessments etc, this may be just too complicated for the majority of projects and could lead to either rejection of processes or, ironically, a decrease in efficiency.
There was confusion over roles. Who was responsible for what? This included the new roles that have recently been spoken about like the PAS1192:2 Information Manager and Task Team Manager. There were discussions about how these roles compared with roles already being adopted such as those proposed by the AEC (UK) BIM Protocol.
No consensus was really reached here, but some key suggestions to understanding these roles were to firstly not treat them as ‘people’.
One person may wear many hats and sometimes more than one person may be required to fulfil one role on a project. PAS1192:2 is a project level information protocol aimed at those used to project management of multi-organisation teams, so the roles may not be a single person or indeed static through the life of a project.
Overall, the project kick-off seems to be where most time was being spent in the development of strategy, and while there is debate there seems to be consensus that the general direction of a more overarching project managed approach was beneficial and starting to work. There seems to be a lot of enthusiasm for what will become possible as a result.
This stage focused on the authoring and delivery of a design Building Information Model. This model would usually contain information from the Architectural, Building Services, Structural, Civil, Infrastructure and Landscaping disciplines, and perhaps the suppliers.
To me it is obvious that if you want to reach the stage of producing the information typically required in an Employer’s Information Requirements document then a co-ordinated approach to modelling is required. So at BIM4Real we asked questions like ‘who owns the stairs?’ They have Architectural, Structural and quite probably Building Services elements.
The idea of the question was to uncover people’s strategy to ensure collaborated or federated models do not end up containing duplicate geometry and of course, duplicated information. There was plenty of discussion, but no consensus as to how this issue should be approached on a project.
Some suggested that the reason for this was software limitation and that cloud solutions would begin to resolve duplication, although I would argue that there must be a strategy nonetheless.
Some suggested this was BIM Maturity Level 3, although that was contested. People did suggest this made a clear case for a detailed Matrix of Responsibilities to be developed.
There was also a suggestion that this showed that Design Review meetings do not always involve the right people; that technical staff with knowledge of how to deliver models should be involved. This would give senior technicians an opportunity to speak with their counterparts from other disciplines and for them to agree how to approach the modelling collaboratively.
It is important that the design leads, contractors, and clients are all in the same room for project design review in order for a collaborative project to work. Software geeks to talk to one another as well.
Validation, a critical part of the handover process of a model to a contractor and also at data exchange points, is largely dealt with through peer checking rather than any kind of formal audit process. Looking at this from a risk perspective it is surely a concern? When we hand over a model to the contractors and fabricators, can they trust it for construction scheduling for example?
Getting design delivery right is key to moving towards a truly collaborative design process. But we need to change the ethos of how we approach it.
There seems to be real momentum now and a good technical and strategic foundation of ability in our industry to realistically push for this.
BIM4Real was a collaboration between Ramboll’s Graham Stewart, Ray Purvis of Atkins and Excitech’s Rob Clark.
The three industry experts were sitting in the keynote presentation at Autodesk University, when they realised that what they wanted from a Building Information Modelling (BIM) event was no longer what BIM was. They wanted evidence of what the industry is doing, what it cannot do, where the real challenges are and what is real best practice.
They thought the best way to achieve this was to get a lot of clever people talking to one another; to bridge the gap between strategy and delivery.
So, BIM4Real was born: a workshop event without any formal presentations.
Attendees included architects, engineers and construction managers, as well as clients and consultants.
Construction and operation
There was a lot of talk about how models are effectively communicated, including the hot topics of OpenBIM via Industry Foundation Class (IFC) and asset registers like COBie. It would be nice to report that there was clear consensus on either of these issues, but it still seems to be a significant sticking point.
It was hotly debated as to whether IFC is the right format for outputting models, with many raising trust issues. However, IFC was seen to be improving in both the understanding of it and what software can do with it; some suggesting that it is up to users to put pressure on software vendors to keep improving this workflow.
I personally have mixed feelings about IFC and how well it moves information from place to place.
If you are happy to really dedicate your time to understanding it, including the various mapping options for data, then you can probably do really well and deliver on expectations.
Most people, however, just want to hit ‘Export’ in one software and ‘Import’ in another and the same model appears.
Without significant expertise that is not guaranteed to happen at present, no matter what software is being used. It is unfortunate that in recent software releases there are countless options for export settings with no real way to test the import until it is sent to the recipient.
As collaborative design becomes more open it would be naive to suggest that a single software vendor will be used on projects, or that vendor formats will be future-proofed for ongoing operation of buildings.
We need to keep pushing our knowledge here and at the same time pushing the vendors to work collaboratively to make these workflows more simple.
The same is often said of COBie. The standardised data concept is fine but the idea of using spreadsheets is questionable. These are a cumbersome format not suited to allowing many people to edit or for handling large quantities of data.
A consensus was formed that interference checking should have occurred prior to handover to the contractor, thus handing the responsibility for this to the Design Lead. However this suggestion caused a lot of argument.
It was suggested that the BIM Execution Plan should make clear the responsibility of who is federating the model and who is conducting interference checking. That is not to say that the Contractor Lead should not continue this exercise as models are updated to reflect on-going site conditions.
There was a suggestion here that the contractor and cost consultant should also be involved in the technical discussion to perhaps ensure that, as far as possible, models were created which are sympathetic to their requirements.
All in all, the BIM4Real round table session did seem to raise more questions than it provided answers.
That said, I think most people felt relieved that they were not being naive in any way by not understanding some of the key issues surrounding the delivery of projects. We, as an industry, should be able to address these issues and work together to resolve them such that we are able to move from the strategic direction to delivering more projects with increasing benefits through BIM.
All of these findings can be downloaded in more detail at the BIM4Real website (www.bim4real.co.uk).
No copyright is claimed; therefore you can use any part of it towards your own efforts.
If you enjoyed this article, subscribe to AEC Magazine for FREE