The COBie Conundrum

I know my posts on LinkedIn cover a very broad range of subjects. You can only imagine what is going through my head every day. However, all my thoughts have the same driver; which is to improve the outcomes and impact of building, ultimately reduce carbon and create inspiring places. That is quite an aspiration.

This week’s thoughts have been around COBie, which may not be considered the most exciting subject and not something I would normally have an opinion on - but bear with me.

I know I’m stepping into dangerous territory as there are many passionate about this area and know a lot more than I do. It also is an area which really gets people mobilized.

My thoughts on all this BIM stuff is that anything we do has to start with ‘why?’ We need to understand the issues and talk to our clients about their issues. We must not fall into the trap of doing things just because they are ‘clever’ or ‘interesting.’

This week I have been drawn into a few issues regarding COBie workbooks with clients. They have had issues with accuracy and how they can view models. During this period, the issues with COBie at a client level have become apparent.

Firstly, it’s great that clients are now recognizing the value of detailed digital information relating to their assets. This is huge progress. However, they are now identifying the shortcomings of the current approach.

I am a big fan of COBie as it has allowed us to structure and align data. When you look at other countries, the lack of this framework has held them back. I’m sure there will be far more sophisticated database technology in the future however, in the meantime, how we do things is generally ahead of others.

The issue is terminology and spreadsheets.

I think some in the sector believe COBie is an Excel spreadsheet. From my experience, the spreadsheet is the issue. Because everyone is familiar with spreadsheets, they are comfortable to input data manually. This human aspect is where things can go wrong. On a recent project, the workbook had been filtered and it sent the data all over the place which frustrated the client as they wanted to cut and paste the room names.

I believe that the best way to do it is to model using the COBie schema and the data should then be kept up-to-date in the model.

Excel is the method of transferring this data to somewhere else. I would suggest that it should never be touched and any change to the data must be able to sync back into the initial model, otherwise we lose the important golden thread.

The syncing of data back into the model seems to be the challenge. It can be done but it is laborious. The default seems to be to ignore the model and populate the spreadsheet to tick the EIR box but how much use is this?

Because everyone is familiar with Excel, they are comfortable to make changes themselves. This breaks links and effectively creates two conflicting databases: the model and the spreadsheet - which one do you keep up-to-date?

Most EIRs request a COBie deliverable which is typically delivered in a spreadsheet . Do we know why?

I can hear everyone shouting to transfer data. I’m very keen to hear examples of where the data has been transferred. I know there are CAFM examples but are there others and how easy and effective has this transfer been?

We seem to have become hooked on delivering Excel spreadsheets filled with COBie without really standing back to think whether this is the best use of our time.

Some companies have developed whole platforms to help fill in this data so that it can be pushed out as a spreadsheet.

Having spoken to quite a few clients this week, what they want is a 3D representation of their building with data accessible at an object level. Nobody I’ve spoken to has found a use for the COBie Excel spreadsheet - apart from the person who wanted the room names transferred into their asset management platform. It seems to be a very complex and expensive way to populate a CAFM system .

The challenge is to ensure the data in a model is accurate and up-to-date. If we are collecting data, it must be pushed into the model and if it has to be transferred, don’t touch the spreadsheet. If you do touch the spreadsheet, make sure it maintains a relationship to link back to the model.

I have now lit the touch paper, I’m sure there will be a lot of response but I’m genuinely keen to gather opinions and hear how my thinking might be wrong. I can then go back to my clients and explain what to do with all of these Excel spreadsheets.