Big data

421 0

 

How can firms take the first steps towards applying big data methodologies in their AECO practices?

If CO2 was the byproduct of the industrial revolution then data is surely the byproduct of the digital revolution. Data flows from nearly everything we do. We produce data when we send emails, take photographs, and check in with Foursquare. We produce data when we use our credit card, the internet and our GPS. And our buildings are producing data about their internal climate, about lighting use, and about room occupation and equipment utilisation.

It is almost impossible to comprehend how much data we are producing. IBM estimates that we collectively produce 2.5 quintillion bytes of data in a single day (IBM 2012).

To put that in perspective, the hard-drive in your computer probably accommodates one terabyte of data, which means we fill an equivalent of 2.3 million of those hard drives with new data every single day.

Perhaps more astonishing than the amount of data is the rate at which our data production is increasing. For all the data that exists in the world — all the data contained within all the books, and emails, and YouTube videos, and wherever else — 90% of this data was generated in the last two years (IBM 2012).

Our data production has been steadily eclipsing itself and with the rate of data production increasing exponentially, in the not too distant future it will be eclipsed again.

This might not sound like architecture. Big data — the name given to all of the data being collected and analysed — has traditionally been associated with marketing agencies, silicon valley startups, and intelligence organisations. But architecture is fundamentally about data.

Advertisement
Advertisement

At the foundation of every BIM model lies data. Architects and engineers work to orchestrate data — creating it, modifying it, using it in simulations. Contracts take data and turn it into physical spaces, while building operators are increasingly managing buildings using data from architects and from building sensors. Data underlies much of the modern AECO (Architecture, Engineering, Construction and Owner-operated) industry and we are only destined to produce more of it.

Data alone does not make big data. In fact, ‘big data’ is somewhat of a misnomer since ‘big data’ does not have to be big.

Big data is really about turning data into knowledge and wisdom; it is about metrics and analysis, insights and decisions. None of this requires a large data set, rather it requires an inquisitive approach and lots of statistical analysis.

Given the amounts of data being produced by the AECO industry, there is a huge opportunity here, but it is one that not many firms are yet pursuing.

At CASE we have been working with a number of our clients help them take the first steps towards applying big data methodologies in their various AECO practices. One of the most common engagements is to develop analytic dashboards that summarise the salient and interesting features in large datasets.

Figure 1 shows a dashboard prototype we are developing for clients that lets them quickly compare design options in terms of environmental performance, leasable area, and program breakdown. This data is all present within the model, but through analysis and visualisation this data becomes the knowledge needed to support design decisions.

With this knowledge designers can parametrically manipulate the buildings with immediate feedback about how their changes will impact the development’s performance.

For other clients we have developed dashboards that reveal space usage in masterplans, that break down post-occupancy usage, that analyse energy requirements, and that apply predictive interaction analysis to office environments.

Beyond analytics for single projects, we have also been investigating the potential of applying big data to collections of projects.

Figure 2 analysis on the utilisation of BIM models within an organisation

Figure 2 shows analysis we have been performing on the utilisation of BIM models within an organisation. This data is often buried in server logs that record how often models are opened, who opens the models, and the model’s size.

These raw usage statistics are surfaced and analysed, transforming the data into knowledge of how the organisation is operating. For another client we have been been data-mining its BIM models, helping to identify common patterns across designs that can be reused on future projects.

We have also been looking at firm-wide performance and turning data about project sizes, location, programme type, and revenue into business development insights.

The data generated by the AECO industry is only getting bigger. Data is flowing from an ever increasing number of sensors embedded in buildings.

There will also be far more data available from things like WiFi triangulation (to track the movement of people through buildings) and satellite imagery (to track the development of cities).

Unlike the CO2 that has spewed from the industrial revolution, the data flowing from the digital revolution is a renewable resource — one whose abundance is increasing exponentially. While few AECO firms have begun utilising big data, those who mange to harness it are set to uncover a rich source of knowledge buried within the byproduct of their digital conversion.

Advertisement