BIM – The end of the beginning?
By Alastair McCapra
The LI’s new book on BIM helps practitioners adapt to the rapid changes that are occurring. But these are only a first step, as the rise of artificial intelligence is set to transform the profession.
Five years ago the LI began its journey of exploration with BIM. Since then a lot of work has gone into ensuring that LI members are well-prepared for the challenges that BIM poses, and the recent publication of BIM for Landscape shares the fruits of that work across the whole profession. What has changed over the last five years is that back then, BIM looked like the big new horizon, whereas now it looks more like only the first stage of something much larger.
In that time we have seen the emergence of driverless cars, wearable technology, contactless payments, and the internet of things. But behind much of this lies a deeper level of technology – artificial intelligence (AI). For the first time, we have machines which can do more than perform specific tasks which they are programmed to do – they can learn as they go along, just as we do. For AI to work, it needs data – big data. By searching quantities of data which it would be impractical for humans to digest, it can find correlations and patterns which would not normally be obvious. It can make life easier for humans by choosing the best route to drive; it can write natural-sounding text or speeches. Potentially it can also evaluate and improve a design.
The BIM process creates a data model of a project, and allows the performance of a design to be measured against its specification. Just as a modern car or aircraft self-monitors and can provide data which is used to generate new designs, a modern building will generate a stream of data about factors such as physical stress, material change, flow of people, and environmental performance. AI can be deployed to learn from this data, and modify the BIM data model to incorporate improvements for future projects.
Up until now the design process in the built environment has made use of various technical tools such as CAD, but these tools only support and facilitate a traditional design process rather than replace it – in much the same way that over the years cars have developed more sophisticated dashboards, but until the advent of driverless cars, the act of driving remained essentially unchanged since the days of the model T Ford. As AI begins to be deployed in the built environment it will assimilate data from BIM models of built designs, combine it with data flows from completed projects, and propose new optimal solutions for the designer to consider in future.
Beyond landscape architecture and the built environment, AI is poised to play a rapidly-expanding role in all kinds of professional contexts – wherever it can offer greater scale, speed or savings over traditional human processes. Use of AI is already particularly advanced in the field of law, where expensive hours of junior staff reading documents to identify issues in a case can already be replaced by e-disclosure, where they are read and annotated by machines1. AI is increasingly being used in the news media as well – a piece of software called Wordsmith2 created more than a billion news stories from raw data in 2014. (The firm which produces the software employs fewer than 40 people).
It is sometimes supposed that the jobs most at risk of automation are those involving repetitive office work, while roles which are more intellectual, more creative or involve the exercise of greater professional judgement are relatively secure. Repetitive office tasks are certainly at risk of automation, but so are many others. In 2013, Carl Benedikt Frey and Michael Osborne produced a paper for the Oxford Martin School on the future of jobs, which concludes with a list of 702 roles ranked in order of likelihood that they will be automated in the near future.3 Those least at risk, in their view, are recreational therapists (1), while those most at risk are telemarketers (702). Landscape architects are ranked at number 133, making them slightly less at risk than mathematicians or editors, and slightly more at risk than vets, writers, or astronomers. In reality, three years after that paper was written, most commentators now think in terms of AI replacing sets of tasks rather than entire jobs, so for professions whose work involves a range of different kinds of task, the question is which of those tasks it will be necessary or worthwhile for humans to continue doing, and which will be done to a satisfactory degree by AI.
I say ‘done to a satisfactory degree’ because another misconception is that as long as humans can produce work of higher quality or greater distinctiveness, they will not be replaced by machines. This is what handloom workers believed during the great industrial revolution, but all experience since then shows that while there is always a niche market for bespoke creations, mass production meets most market needs. For most clients in any market, quality and distinctiveness are usually lesser considerations than speedy delivery, basic serviceability and price.
The rise of AI poses all kinds of new challenges to the professions. For one thing, the value of accumulated human knowledge and experience will fall, enabling relatively sophisticated tasks to be performed by more junior staff. For another, earlier generations of software such as Sage Accounts or AutoCAD were bought by the professional for use in their office. Emerging AI systems like Wordsmith and Cyfor are however provided as an outsourced service. This distinction could have profound implications for professionals’ self-perception and status, because if they use outsourced AI services, they are transformed into consumers of the services they used to produce. Likewise, the ability of the current generation of professionals to provide expert high-level analysis and interpretation derives, at least in part, from their years of experience in sorting, processing and analysing the material that supports key decisions. In the future, AI is likely to replace humans in these processing roles. How then is this expertise to be built?
BIM has evolved primarily to improve engineering processes, and has developed to encompass first architecture, and then landscape architecture. In contrast, AI did not develop in response to specific professional challenges, but it is beginning to have an impact on a number of them, including in architecture and urban planning4 5 6. When BIM and AI intersect fully over the next few years, the result will be transformational for all disciplines in the built environment. We can imagine a future in which large construction firms have live data models of dozens of built schemes of different types, and use them effectively as a menu of oven-ready designs which can be modified as necessary to suit client requirements. The role of professionals working with such models will be to understand and negotiate variations to them, based on the constraints of the site and the requirements of the client. There will certainly be some clients that want a completely new design for a building, but in the future commissioning a building will be much more like buying an aircraft. There are a limited number of basic designs available from manufacturers, and while client needs can be accommodated, it makes sense for most to buy off the shelf.
What does all this mean for the future of landscape architecture? Compared with professional activities which are more purely conceptual or analytical, landscape architecture will probably experience change more slowly and partially, so there will be more time to adapt. My guess is that many tasks related to planning, including much of LVIA, could be largely automated in the future, as Site Information Modelling (SIM) develops alongside BIM, and AI helps identify optimal solutions for the siting of turbines, housing developments or other schemes. At the same time, if the core structure of buildings is going to become more standardised, the role of landscape architecture in ensuring that our built environment does not become uniform and characterless will probably be more important than ever. The critical importance of management and maintenance of completed schemes will perhaps come more clearly into relief, and there may be new tools to help monitor them and highlight the need for remedial intervention. Above all, community engagement work cannot readily be automated, so whatever new technical aids may be deployed, including augmented reality experiences, direct human involvement by experienced professionals will continue to be required in order for successful schemes to be delivered.
BIM effectively opens the door to this brave new world, and ‘BIM for Landscape’ is an essential primer in the methods and technologies which will be used to capture, record and interpret data. Over the next decade, working methods within professions and collaborations between them will change in extraordinary new ways as AI is more widely deployed. It is good that the Landscape Institute is placing itself in the lead for the interesting times ahead.
Alastair McCapra is chief executive of the Chartered Institute of Public Relations.
He is a former chief executive of the Landscape Institute.
BIM for Landscape shows how BIM can enhance collaboration with other professionals and clients, streamline information processes, improve decision-making and deliver well-designed landscape projects that are right first time, on schedule and on budget.
This book looks at the organisational, technological and professional practice implications of BIM adoption. It discusses in detail the standards, structures and information processes that form BIM Level 2-compliant workflows, highlighting the role of the landscape professional within the new ways of working that BIM entails. It also looks in depth at the digital tools used in BIM projects, emphasising the ‘information’ in Building Information Modelling, and the possibilities that data-rich models offer in landscape design, maintenance and management.
BIM for Landscape is published by Routledge and costs £39.99.
BIM for Landscape was produced with the help and support of the BIM Working Group who steered the project from the beginning. Many companies supported individual BIM Working Group members; and the LI IS grateful to registered practice Colour who were commissioned to write the book and supported and managed Henry Fenby-Taylor to prepare and deliver it.