The trouble with BIM

With the UK Government now mandating the use of Building Information Modelling (BIM) on its future projects, the AEC industry is facing sweeping changes in long established processes and collaboration practices. Martyn Day looks at the challenges of moving to BIM.

Building Information Modelling (BIM) has had a lot of hype over the years, yet the industry has been slow to change or adopt this new technology. It is estimated that as few as 10 percent of firms in the UK construction industry use BIM software on their projects. This compares to approximately 60 percent in the USA.

While those in the private sector still have complete control over their choice of design methodology, the decision to enforce BIM on government infrastructure and building projects will force many firms to invest in new training and software. Without this they will not be included in future government contracts. This is without doubt an industry transformative decision.

The drive comes from Paul Morrell, who is the Government’s chief construction adviser. He stated that he was “convinced that this [BIM] is the way to unlock new ways of working that will reduce cost and add long-term value to the development and management of built assets in the public sector.”

In many ways I agree with Mr Morrell that the theoretical benefits that BIM brings, will finally address the efficiencies that Egan and Latham identified all those years ago in their renowned government reports. However, the problem is going to be mapping out how to get there, how much it will cost and how long it will take to get the return on the investment. Fortunately there is some breathing space, as the final compliance date is 2016.

When Mr Morrell was asked what he thought about architects not being able to afford the software upgrade and training, he replied that he was more worried about the cost to practices that do not make the move. He added that firms that stay outside the BIM methodology would “feel as disconnected as one without email”.

It is clear from reading his other comments in the mainstream architectural and construction magazines that Mr Morrell is expecting the whole spectrum of UK AEC professions to eventually adopt BIM for both private and government contracts, as it will be proven to be the most efficient way of working.

Drive to change

While I like the vision and the commitment to this drive to improve efficiency in the UK construction market, I am worried about where we are now as an industry versus where Mr Morrell wants the industry to be in five years, especially considering the capabilities of the current BIM software, the average user’s skill-set and the general lack of experience in completing 3D BIM projects in the UK.

The other dynamic at play here is the ingrained culture of working and the contract structures that do not reflect the shared risk/rewards that a truly collaborative project would require. BIM is not just about buying some software and training, it is about learning to work with other project participants differently and striving to make efficiencies.

Even firms that have made the move to BIM have found that as many as three or four projects need to be completed before the process can be internally ironed out and the benefits of BIM realised. Managers need to be prepared for less than instant results and occasionally have their faith tested.

I regularly talk with firms that have adopted BIM and the reality is that while their belief is unshaken that it is undoubtedly a better way of working, offering benefits, there are some ongoing headaches. While 2D CAD has its problems, and BIM solves some of these, it also introduces some new issues of its own that need some creative solutions.

So let’s look at some of the myths and common problems that can be found in the BIM world.

BIM utopia

BIM is presented as a one stop solution: build a single, data-rich model containing architectural, MEP and structural elements and all the drawings will be produced automatically — plans, sections, elevations, the quantities, the renderings, the costings and even perhaps the energy certification. This is a utopian vision compared to what current systems are truly capable of and neglects to describe how BIM front-loads the design process with many decisions that would normally be made late in the process, brought forward.

The richness of BIM is in the Information. The model can produce parts list schedules and quantities to help provide more accurate costings earlier in the design process. Image: Autodesk Revit

With BIM, the workload changes and shifts forward many design decisions. Before starting a project, new ‘families of parts’ (intelligent components) should be considered and created, which may bring forward decisions on interior fittings such as doors and windows, escalators, lifts and glazing. This can be beneficial, using analysis tools to refine the performance of the design — but then again these activities are also brought forward.

The whole single building model (SBM) is really only available in a dream. An architectural BIM model is really quite different from a construction BIM model — while the geometry may be the same, the information contained is quite different. The construction scheduling needs elements within the model to be cut down (e.g. how many concrete pours for the floor), which will not be included in the architectural BIM model’s representation of the concrete elements. Companies like Tekla creates models for constructability and fabrication with the corresponding detail. Vico is another firm that creates BIM models for quantity, cost, scheduling and production control.

The single model utopia also does not make sense from a hardware perspective. Depending on the software being used, BIM models get very big, very fast. David Light from HOK recently told me that the firm is stipulating 24GB of RAM for its new workstations so it can hold the Revit datasets without paging to disk on its biggest projects. This is a considerable investment. BIM simply cannot run on old machines designed for running AutoCAD or VectorWorks. Other BIM modellers are not quite so memory hungry but they still quickly start to swell as the model detail grows.

Autodesk’s HQ building in Boston was modelled in Revit when the firm renovated the interiors. This was an impressive model and allowed Autodesk to practice what it preached. The one thing I could not understand was why clash detection on the project was not done in Revit but in Navisworks Manage, which is a considerable additional price.

I met one of the architects who told me that to perform a clash detection, the geometry of all the disciplines needed to be loaded and this made the Revit model so large that only one of Autodesk’s machines was capable of loading the complete model, let alone running clash detection.

Navisworks has the advantage of loading a lightweight version of the model, mainly the faceted geometry without all the memory-hungry BIM data. Navisworks still has enough information to perform an intelligent clash detection of a complete BIM model.

While there are advances in hardware approximately every nine months, using BIM is going to drive an appetite for hardware and users will most certainly need a 64-bit operating system, lots of RAM and the best processors they can afford. The bigger the project the more likely it will need a strategy for cutting up the model into manageable sections.

The level of detail in the model is also highly important — too little and the benefits of modelling would be lost, too much detail and the model will become too big to load and manipulate. In the early days, many BIM implementation failures were because firms added too much detail to their BIM models leading to unwieldy file sizes.

Accept the output

The early BIM systems were mainly sold on the productivity benefit that could be had from getting automatic output of 2D plans, sections and elevations from the 3D model. The view that the 3D model was the source of all the content simplified the process as there was one version of the truth and traditional 2D CAD was susceptible to errors from lack of co-ordination.

While correct, the output from BIM systems is not necessarily compliant to the myriad of individual styles that firms like. While they can be configured to a degree, many firms take the drawings and then feed them through AutoCAD or AutoCAD LT to comply.

At this point, they introduce a disconnect between the BIM system and the drawings, which could introduce the same co-ordination problems they were trying to avoid in the first place. Any changes to the model, will change the drawings and these will all need to be hand edited, again.

Today’s BIM products can produce great 2D output, mixed with stunning 3D rendered visuals. It iss best practice to accept the styling capabilities of the BIM system and discourage external editing.

A BIM ideal: Structural, MEP and architectural in a single Bentley MicroStation model. Takanaka Corporation used BIM for design and construction simulation to build the ten floor building. The company estimates it saved 10 percent in design hours, reduced construction workers by 15 percent and reduced construction time from 14.5 months to 13. Image courtesy of Takanaka Corporation.

Data exchange

Data exchange has always been something the CAD industry has been terrible at. With BIM, it appears that the competing software players are taking interoperability to new lows. 2D CAD was just lines, circles and arcs and look at the problems that were generated simply by exchanging DWGs.

BIM models have complex 3D geometry, behaviour, parametrics and lots of information attached to each object. Data translation between BIM systems is an order of magnitude greater than the previous generation of CAD.

The only real independent standard available is IFC (Industry Foundation Classes). This format is a documented standard, which the vendors unfortunately interpret and implement at their leisure. The end results are hit and miss, models are always dumbed down from the original and even when using IFCs to send data between products from the same vendor, you can not guarantee success. However, IFCs are the only real hope and one can only wish that this improves. For now a number still share data through 2D drawings (DWG and PDF).

Here, I think Paul Morrell has missed a chance to get the key BIM software players in a room and bash their heads together to improve data interoperability before handing them the UK’s AEC market on a plate. As things stand, Morrell makes reference to COBie, which is the Construction Operations Building Information Exchange — an Excel spreadsheet data format that contains non-geometrical information about a building and was developed in the US. COBie is no magic bullet, if defines floors, spaces, zones, components and attributes. I assume that the geometry will be from some other format, probably IFC.

For now, it seems that information held within multiple BIM models will require export in multiple ways to get hopefully reassembled in a different system, not unlike Star Trek’s transporter.

Unfortunately one of the benefits targeted by Mr Morrell is the need to collaborate better. The poor data interoperability of today’s BIM systems is pushing this benefit in the opposite direction. The solution is for everyone working on a project to use the same BIM system, which will lead to either firms owning multiple BIM tools, depending on who they are working with or the dominance of one vendor’s application throughout the country.

Cultural issues

Changing a long established way of working is always going to cause some problems and BIM has its own unique characteristics that need addressing. Obviously there is the ‘people’ issue, staff skills need to be upgraded and new staff will need to have BIM understanding. Internal processes will change and new teams built. Most start with a pilot project to learn and then work their way to bigger projects.

While BIM tools are ‘easy-in’, there is actually a fairly steep learning curve once beyond the honeymoon period. It is important to identify internal BIM champions who will understand the technology and be there to help the rest through the transition.

BIM requires internal standards. To be honest, 2D CAD also required internal standards and many failed to get organised in 2D. Due to the complexity and way of working, BIM cannot be effectively used in a collaborative environment without defined standards. There is a workgroup that has defined the AEC (UK) BIM standard for Revit users, which is a template that can be downloaded and incorporated into protocols. The standard has no legislative backing, it is written in the context of rules to be followed and is aligned with BS 1192:2007. http://aecuk.wordpress.com/.

True collaborative working with shared models alters the very nature of the existing contract structure. Firms have to rethink the way they work with each other, dumping the silo mentality and thinking more of integration and connected project teams. This is a cultural shift that will probably go so far as to actually alter the structure of most AEC firms. BIM actually favours design/builders, those that already have integrated teams. The traditional grouping of firms to complete projects in a federated structure requires a new attitude to liability of deliverables and an approach which shred the risk and reward. In the US this is decribed as IPD (Integrated Project Delivery) with everyone in it together. As we work a little differently here, the UK will have to come up with its own alternative.

Finding like minds

Finally, when using BIM it pays to work with firms of a similar mindset. With the interoperability issue, it helps to be on the same system — Revit Architecture with Revit MEP or Bentley Architecture with Bentley Structural.

As BIM is still in its formative stages, adoption is fairly thin on the ground, especially within the MEP industry. Working with multiple systems and methodologies will not deliver the productivity benefits originally claimed until adoption accelerates.

Conclusion

Building Information Modelling is widely seen as the next technology for the building industry and rightfully so. However, the technology was not written for the way the industry works today or has worked in the past. Adoption will require adaption, investment and greater teamwork between the core trades.

There are still a variety of complex issues that need to be overcome both in the technology and between the developers of these products. Those that are using the tools and seeing benefits, are slowly finding solutions to their problems but realise that the software firms are continually improving their products’ capabilities to cure the headaches. However, a lot of effort still needs to be applied to helping data move more seamlessly between competitive systems and all adopters should realise that the BIM utopia is far from being delivered on.

The UK government has set the BIM ball rolling and now many firms have a deadline in which to comply. Fear of the unknown is not really an option for those using traditional 2D CAD. Fortunately there are a number of events and industry experts publicising initiatives to help the UK’s construction firms get up to speed on what this new directive will mean.