How to use your data to take advantage of the digital revolution

From the mundane to the magnificent

Data and making it flow will be central to the next paradigm shift in the way we design buildings, which will be much more transformative and disruptive than the transition from drawing board to CAD.

Data is in many respects counterintuitive for most architects: we are primarily driven by shaping geo­metry and a building creates its greatest value by transforming communities and cities, enlightening and inspiring us to perform better wherever we are: school, university or our workplace, be that the traditional city centre office or, in the new model, in our homes or neighbourhood co-working space. However, this is a game changing topic and better data may significantly improve performance outcomes over the whole life of a building. For example, data from sensors and Internet of Things (IoT) technologies moves beyond classroom utilisation or room temperatures towards understanding how lighting, temperature, oxygen levels and humidity affect our performance throughout the day to allow the ultimate goal of real-time adjustments and optimisation of our environments.

The earliest BIM innovation came from connecting geometry to data when Frank Gehry used aerospace drafting technology to define the curves of his buildings for contractors. It was data for buildability. Architects seized on the possibilities, leveraging parametric modelling (where geometry is connected to visual scripting tools) to allow dynamic new forms to be quickly adjusted and adapted. This is now commonplace, for sports stadia in particular. However, these examples only underline data’s subservience to geometry.

Models were devoid of other sorts of data but it is the wealth of it elsewhere in the design process, such as specification data or the space data informing engineering analysis, that unlocks design workflow innovation. Before we can harness the power of these datasets we need to start, as a profession, looking at how our building information models are structured, beyond the data-lite, traditional scaled drawing containers that are pushed out of them. Innovation starts with the simplest of topics: data classification. Understanding how to structure our data is absolutely crucial. By using Uniclass 2015 to classify spaces, systems and products we set our content up for the long term, making it machine readable for an increasing array of artificial intelligence initiatives.

Grappling with complexity

We live in a world of change; intuitive knowledge decreases in value as more new complex topics such as the circular economy or offsite manufacturing demand a reset. As we rebuild our knowledge around these topics, data will redefine the profession by enabling the lead designer to make better decisions and deliver more informed insights on projects large and small, with less reliance on an individual’s knowledge as we turn to industrywide research and feedback from completed projects. Industry can only hit ambitious carbon reduction targets if we share experiences. The planet can’t wait for everyone to rebuild their heuristic knowledge. We can all have our own brand differentiators whilst simultaneously driving progress towards net zero carbon solutions.

Unlocking real-time collaboration

Multidisciplinary working reliant on rules of thumb generates a time lag between creating design solutions and receiving engineering feedback, requiring multiple iterations of the design as information flows back and forth and is progressively fixed. This is not only inefficient, it allows sub-optimal solutions to be given too much airtime. Imagine a transformed process with high quality data at our fingertips. For example, having real-time data on energy, carbon, cost and daylight will allow immediate honing of facade options. Iterations will peel away, making way for real-time decision making.

Working with our engineering collaborators, we must use our creative skills to imagine the data journeys that would transform our designs, particularly in terms of sustainability. Design automation where connected data flows frictionless from one spreadsheet, or software package, to another, is not about standardisation nor commoditising what we do, but a design process based on facts and greater consistency from one project to the next. Think of stage 3 being automated inside stage 2. Engineering analysis considered in real-time as part of the optioneering process.

Drawings to libraries

We must also move away from the focus on general arrangement drawings. The future will see libraries of construction and/or manufacture ready spaces full of data assembled in profoundly different ways. For example, libraries of operating theatres, imaging suites and consulting rooms allowing us to engage with healthcare clients using photo-realistic, VR information from the outset.

To many, algorithms sound as though they come from a different world but at a simplistic level they are just about following rules and logic. Overlaid on to each practice’s ‘data story­telling’ they can be used to speed up the design process, avoiding decisions needing to be made repeatedly from one project to the next, checking our models against regulatory or sector requirements.

The future is now

While this can feel a long way off, software companies are already targeting the architect's building information model, developing plug-ins that will provide a broad range of simulations faster. They can steer us more confidently towards a solution, for example by determining the right energy or fire strategy earlier in the design process.

Generative design is coming centre stage in 2021, as a way of crunching thousands of options against a set of variable design para­meters. Although some, including Make and Sidewalk Labs, are creating their own version of these tools, industry-wide ones are becoming available to all, allowing any practice to use human knowledge to set the parameters and their intuition to distil thousands of automated machine options into the most viable and relevant ones.

First steps

The ability to fully leverage the power and potential of data will not develop overnight. Artificial intelligence is a long way from being business as usual. But by starting now, practices can incrementally build their skills. This might mean making sure models are classified correctly or using free tools such as Prism. But it can also start at a more fundamental level by imagining what data would transform your design process; perhaps real-time data parsed from completed projects or how data connected across the design team, such as carbon information, might result in a faster and more effective design process. Let’s all begin the journey today, to create an exciting future tomorrow.

Dale Sinclair is director of innovation at Aecom