Analysis

COBie inventor Bill East: ‘time to put an end to bad data’

The widespread misunderstanding of Construction Operations Building information exchange, better known as COBie, in the UK is a direct result of the government’s top-down approach, says Bill East, its inventor. In the first half of a two-part interview, he talks to Denise Chevin about the COBie reboot and the challenges that lie ahead for digital asset handover.

When you first invented COBie in 2007 did you envision it becoming such an industry standard for the digital transfer of asset information?

Having spent my early career as a project engineer, it was obvious that critical building information data had to be manually re-discovered and re-typed many times. That is about as far away from efficient as could be imagined.

Twenty years after my project engineer days, when the problem of building construction handover was still not resolved, I jumped at the opportunity to help a colleague from NASA try to do something about it. 

In the original 2007 report, I noted the need for COBie’s international adoption. The validity of that recommendation could immediately be seen in worldwide recognition of the picture of the maintenance manager in a boiler room in front of boxes containing handover documents. The name of the person in that picture is Lyle, not “COBie Carl” as is sometimes called in the UK.

The UK top-down approach meant that the government-sponsored project was required to respond to some hard questions. The result was a nationally mandated specification that was not possible to understand or implement.– Bill East

How did COBie come to be used globally?

The initial 2007 specification was a process-based description of who needs to collect what information during the construction administration process. However, software companies asked for a normalised schema to simplify their implementation. Software companies also wanted to have a single international specification. As a result, in 2009 BuildingSmart international published the Basic Facility Management Handover Model View Definition.

To make the set of information required understandable to non-programmers, I also continued (over the protests of some involved) to insist that non-geometric data could be presented in spreadsheets as well as in BIM models. Why? Because most of the information needed for construction handover is not generated in design authoring software but in the construction back office. The BIM model, for the data required by COBie, simply provides a fancy user interface, but many other less expensive interfaces are also possible.

An updated spreadsheet format that did not fully conform to the 2009 BuildingSmart International definition was taken up as the basis for the US COBie specification. That same specification was also referenced in BS 1192:4. Unfortunately, the UK top-down approach meant that the government-sponsored project was required to respond to some hard questions. The response was not: “Let’s do buildings first since we have good tools for that, and we’ll do infrastructure later once those tools are developed.” The response was: “We’ll find a way to shove all the requirements in there no matter what.”

The result was a nationally mandated specification that was not possible to understand or implement. There were many consultants selling services, often to owners, who judging from the feedback I receive across the Atlantic, know little if anything about the actual goals or requirements of the original BuildingSmart international definition or the US specification.

There have been several examples of government mandates from many countries imposed because bureaucrats decide that if they say it must be done, it will be done. By the time we get to the construction trailer however, if it does not make practical sense, the requirement is simply ignored.

Can this get better?

BuildingSmart International has published a lesson learned report documenting a decade’s use of the US/UK national specifications. Based on these findings, it has begun a project to update their original 2009 standard.

This updated standard removes much of the complexity of the prior national specifications by re-focusing on the original core requirement – delivering the list of equipment installed in a building along with the required maintenance requirements. To get a sense of the bloat in US/UK specifications, the updated BuildingSmart International standard is currently slated to remove about a third of the data elements that were specified, but never used, in the US and UK standard.

To emphasise the focus on the core requirements and to better explain how the complete set of facility management requirements may be addressed, the name of the BuildingSmart International updated project is “Facility Management Handover – Equipment Maintenance Model View Definition.”

Allowing non-tested software to be used is not an issue with the specification, it is an issue with the incorrect application of the specification. This problem was compounded by people trying to force the delivery of COBie data for building elements and project types for which it was never intended.– Bill East

The idea is that there is data required during FM that needs to be defined for different use cases. Examples include equipment maintenance, but also janitorial services, key schedules, painting contracts, etc… A more complete list can be found in the BuildingSmart International Lessons Learned report. The “Handover" part of the name defines what part of that data contractors must hand over to boot-strap the facility management process.

Having this type of framework means that once we have completed the Equipment Maintenance Standard, other standards can be quickly agreed upon under a common set-up.

There has been criticism in the UK in terms of it not offering the interoperability that some would have hoped before between different BIM software: what are your thoughts on that?

Ensuring software vendors complied with the COBie specification was an essential part of the process. In the first decade of use, COBie was tested in more than 30 commercial off-the-shelf software systems in the US. Allowing non-tested software to be used is not an issue with the specification, it is an issue with the incorrect application of the specification.

This incorrect application problem was compounded by people trying to force the delivery of COBie data for building elements and project types for which it was never intended. Using non-standard applications of the COBie specification even within tested software is also a recipe for failure. As before, this is not an issue with the COBie specification, but a failure to understand the purpose of the standard. In a way, it is like trying to use a hammer to drive a screw. It might work, but you will likely get hurt somewhere along the way. 

You recently said that the UK-variant of the COBie specification was partly responsible for the difficulty in implementation: how confident are you that this new standard will iron out any problems?

In my view, the widespread misunderstanding of COBie in the UK is a direct result of the government’s top-down approach taken in COBie’s implementation. Most people will have heard something about this in the context of BIM Level 2. Mandating something that no one knows how to do set the UK up for failure from the start. The need for a soft landings effort can only be seen as the proof of that failure. Adding to the confusion were “pop-up experts” providing inaccurate information to implementers, most of whom, according to the anecdotal evidence I heard about, did not even take the time to download and read the actual specification.

If top-down national industrial policy is undertaken without testing, professional certification, education, and stakeholder outreach, those working on projects will never have the capacity to meet any such policy, COBie or otherwise.

Read part two: https://www.bimplus.co.uk/analysis/cobie-inventor-bill-east-time-put-end-bad-part2/.

Image: 130701064 © Stepanenko Oksana | Dreamstime.com

Story for BIM+? Get in touch via email: bimplus@atompublishing.co.uk

Latest articles in Analysis