The world is on fire. Our governments are broke. Disinformation is rife. At the intersections of these challenges: data and the built environment. Society needs a built environment that meets its needs equitably and sustainably, one that rapidly atones for the sins of the past. Ahead of his appearance at Digital Construction Week, Ian Gordon asks: will data get us there? Or is it simply an expensive distraction?
I’ve been mulling it over. Perhaps we all have. Train delays, power cuts, potholes, flaky broadband. We pause, and we wonder: why? There is so much accumulated infrastructure, seemingly enough to meet our every need. But it sure doesn’t take much for the cracks to show, literally or metaphorically.
In our lifetimes, we have seen intangible assets become the dominant asset of our economy. Our consumer products allow us to generate and use staggering amounts of data. AI has gone from an eccentric distraction to borderline miraculous. Yet our experience of applying these rapidly maturing technologies to the built environment is underwhelming.
The complexities of our industry lie not in the technology, but in people, professions, processes, culture and the primacy of physical assets. Digital technology may have transformed music, socialising, news, entertainment and romance into ubiquitous and weightless things. And yet bridges, buildings, pipes and cables today aren’t far removed from bridges, buildings, pipes and cables from decades ago. No matter what version of the iPhone we carry in our pockets, we humans still have the same basic needs of shelter, food, water and warmth.
“Data is not and cannot be neutral. We must use it to help the built environment meet this larger purpose.”
After a year of reflection, reading and writing on this topic, I must admit that I haven’t quite found the pithy meta-narrative that I sought to describe data’s role in our industry. In lieu of a snappy tagline, the best that I can offer are diagnoses. In isolation they are not so dissimilar from the challenges faced by other industries. Combined, they describe the unique combination of history, culture and technological limitations that make applying data to the built environment a unique and potentially profound task.
The built environment exists to meet societies’ needs. Publications such as the National Infrastructure Commission’s Data for the public good, Centre for Digital Built Britains Flourishing Systems and the government’s Transforming Infrastructure Performance: Roadmap to 2030 remind us that we should be seeking nothing less than public wellbeing, resilience, equity and sustainability. In this context, data is not and cannot be neutral. We must use it to help the built environment meet this larger purpose.
There is an opportunity cost to all investment in the built environment, one measured in carbon. The direct and indirect impact of the built environment on global emissions is enormous and under-acknowledged. Everyone knows that plastic straws are bad for the environment, but fewer people are aware of the disproportionate footprint of cement. Sustainable solutions do exist, and data has a vital role to play in keeping score, making the industry’s carbon footprint more visible and reducing wasted resources.
To work with the built environment is to reckon with complexity. Complexity rarely stems from the assets or the technology, but from the human, societal and political context. Data solutions that do not reckon with this complexity are unlikely to realise lasting change.
“If our digital domain bears no resemblance to our physical domain, then we are deluding ourselves.”
The built environment relies on an incredibly broad range of data: structured alphanumeric, documents, models, maps, imagery and datasets that vary in scale and importance. Deriving value from data on the built environment demands a more comprehensive approach for which there is not always an off-the-shelf solution.
To realise value, our digital domain must act in aid of the physical domain. It’s cheaper and safer to make mistakes in models than in real life. However, our decisions are only as good as the connections between the two. If our digital domain bears no resemblance to our physical domain, then we are deluding ourselves.
Tribalism and specialism
The industry has accumulated a wide range of professions that trace their origins back to Victorian old boys’ clubs. These professions cross our organisational boundaries and we have built these silos into our culture. Data is our opportunity to create shared meaning, common definitions and interoperability to begin to optimise across rather than within silos.
Bias towards building
Creating and improving our built environment is what inspires us, we see the built environment as a morally good thing. Delivering projects is also what brings money into our organisations and what gets us promoted or rewarded. And yet, as we have seen repeatedly, there is an opportunity cost of building in terms of money and carbon, and the numbers do not always stack up.
“Data is our opportunity to create shared meaning, common definitions and interoperability to begin to optimise across rather than within silos.”
We build, maintain and operate the built environment through the coordinated labour of human actors. This means that our success in applying data to the built environment relies upon our ability to improve the flow of knowledge between individuals and computers.
Keeping these diagnoses in mind has helped me to think about how the built environment can benefit from emerging data-intensive technologies. Take Generative AI (ChatGPT, Bard, Midjourney). While it is obvious that massive neural networks now have profound capabilities, it is less obvious how those capabilities can help us to better serve society or reduce emissions.
The diagnoses above start to offer clues. We know that the built environment comes with a data footprint of material spanning hundreds of years. This is far more information than any one human could ever feasibly internalise. But it is a trivial amount of information for generative AI. We now have access to a technology that can synthesise our entire corpus and explain it back to us as usable tacit knowledge. In the past we relied on graphs and dashboards, now we can just chat!
Generative AI has the potential to quickly integrate and cascade knowledge across the industry. This connection of digital and physical domains allows AI to make sense of information coming from the physical world, whether that is real-time IoT feeds or reports submitted by specialists. This promises a decision-making (or decision-augmenting) engine that can reckon with the systemic complexity of our built environment.
The history of blindly assigning decision-making to inscrutable AI isn’t great. By assessing this technology in a manner that is mindful of the built environment’s idiosyncrasies, we are more likely to arrive at solutions that deliver value.
Investing in data
Investing in data has the potential to transform the way we approach the built environment and meet the ever-evolving needs of society. However, this can only be achieved if those using it remain aware of the unique challenges posed by the built environment. Data provides us with a comprehensive understanding of the impact of our work – both positive and negative – allowing us to be accountable and make decisions that benefit society and the natural world, rather than solely focusing on personal gain or career advancement.
With data, we gain a broader perspective, enabling us to recognise our biases and explore new ways of building, maintaining and operating. Data will have an impact on our industry regardless – our meta-narrative should be ensuring that its impact is positive not corrosive.
Ian Gordon is head of data, restoration and renewal at the Houses of Parliament. You can hear him warm to his theme at Digital Construction Week on 17 May.
Don’t miss out on BIM and digital construction news: sign up to receive the BIMplus newsletter.