Sunday, January 9, 2022

Coming Fast: A Driver of Industry Change



Unfailingly, I’m drawn to news articles and/or essays about technological advancements within the architecture profession. I recently happened to read two especially thought-provoking pieces online about the future of BIM (Building Information Modeling). One of these is by Bill Allen, CEO and president of EvolveLAB LLC on the Autodesk University website. He posits the future of BIM will not be BIM at all but rather a new technology—Building Information Optimization. The second is a two-part series by blogger Graham McKay on his Misfits’ Architecture platform (with input from Mohammad Saad Ahmad) that questions whether BIM is driving the industry, or the industry is driving BIM. Together, these blog posts underscore the accelerating pace of technological change in how we design and construct buildings and the implications of that change for architects.
 
The Future of BIM will not be BIM and it’s Coming Faster than You Think
In a nutshell, Bill Allen says our current design and construction processes are poised for significant change, with increasing reliance upon algorithmic thinking and ultimately artificial intelligence (AI). Rather than manually modeling walls, doors, and columns in our BIM programs, Allen foresees feeding computers “rules” that tell them how to develop a buildings’ optimal footprint, structural load capacity, thermal performance, etc. He argues (correctly) that most architects currently operate BIM software inefficiently, at worst using it like a CAD program rather than capitalizing on the vast potential inherent in using data to inform the design process. One example he cites of how architects might exploit the full potential of BIM is to optimize interoperability between the design tools they already use, such that information is automatically generated in one piece of software and translated to another (i.e. capturing the manipulation of a study model developed in SketchUp in real time within a linked Revit file).
 
Allen describes how developments in AI will progressively automate numerous design processes that are presently manually performed tasks. Among the most exciting of these is a soon-to-be available generative design tool that automatically routes ductwork and piping while being object-aware and avoiding conflicts with the building structure and other MEP elements. This AI algorithm will save enormous amounts of time that otherwise would require laborious coordination by the project team during the design process (regardless of the current availability of numerous clash detection add-ins for IFC data models).
 
I’ve long envisioned AI-based tools that would automate code analyses for land use requirements, fire & life safety regulations, energy conservation codes, and accessibility standards. These tools would automatically generate code-compliant designs. Ensuring conformance with these (and other) disparate sets of rules is always a daunting task, so I welcome any advancements in AI that would automate and streamline code evaluation processes.
 
The American Institute of Architects says its members should welcome artificial intelligence as a tool to augment practice and replace mundane tasks rather than as a threat to their jobs. I agree. The x-factor is the complexity inherent in the design process and making of architecture. I recently argued that architects bring enormous value to projects by virtue of their ability to oversee the integration of the full range of design considerations. This will not change any time soon. Computers cannot yet fully connect and synthesize multiple and complex fields of knowledge (especially those that resist quantification) as architects must do on every project.
 
AI cannot engage in the level of multidisciplinary thinking necessary to arrive at the most cost-effective, sustainable, and attractive design solution to each unique design problem. For the near future at least, the level of analysis enabled by computer technology will not approach that which the human mind is instinctively capable of, nor will a limited set of primitive algorithms autonomously generate truly profound works of architecture.(1)  What AI currently lacks is the power of creativity, a general feature of human intelligence.
 
The Future of BIM as an Active Driver of Industry Change
In his two-part essay, McKay asserts the demands of managing a cloud-based BIM model for large, complex projects are driving a trend toward control of the model by construction managers and general contractors, rather than architects:
 
“We’ve all watched BIM become mainstream and our relationship to it evolve but, even until relatively recently, it was still possible for very substantial and complex buildings to be conceived, designed, and constructed without algorithms to generate their geometry, visualizations to picture them, CAD to document them, and BIM to coordinate their construction.

". . . roles within the building design and construction industry have changed to streamline the process but with competition for control of the BIM model. This is not control for control’s sake but can be justified in terms of higher quality, quicker response to change, and a corresponding reduction in the design and construction phases. All these are good things but owning and controlling the BIM model is where the most efficiencies are to be gained and the greatest profits to be made. It is increasingly common for the BIM model to be controlled by the contractor, and other roles subsumed beneath it, that of architects included.”
 
According to McKay, architects can’t be the “traffic controllers” because we typically lack the requisite engineering, construction, and management knowledge. That said, McKay also believes there is no reason why architects can’t become masters of information systems management and design to performance-based outputs and digitally managed information. The problem is attaining this mastery is presently beyond the reach of most small firms. Small (and medium-sized) offices like mine simply lack the assets large corporate firms have at their disposal. We can’t justify the costs associated with training and dedicating staff to this role, nor can we afford to bring on the highly specialized expertise necessary to coordinate, trouble-shoot, and fully capitalize upon BIM.   
 
Regardless of the challenges posed by the need to manage the exponential burgeoning of information and data sets associated with a typical project, there’s no doubting the benefits of BIM for even the smallest of architectural offices. BIM is based on several principles that can be implemented with relative independence. These include assembling a virtual, 3D model of a future building, which is continuously modified throughout the project’s life, valuably even after the completed building is in use. The model includes attached metadata, such as the material properties of a building component, and parametric modifiers like the height of a wall. Preparation of specifications can be tied to the metadata embedded in the model, and objects linked to online libraries can be updated automatically. And of course, because the BIM model is an electronic simulacrum of the real thing, immersive visualization of an unbuilt design is possible.
 
Despite McKay’s perception that architects may be at a disadvantage across the BIM landscape, I’m confident AI will eventually simplify and alleviate the burdens of information management while lowering costs. As a result, architects stand to regain a measure of the control and influence ceded to construction managers and contractors in recent decades in response to the increased demands and complexity of projects. As McKay avers, recognizing design consequences (cost, embodied energy, energy performance, code compliance, etc.) in the form of immediate and actionable feedback during all stages of design promises to minimize inherently wasteful, serial design “refinements.” The return on investment in BIM and AI will thus be considerable.  

Change is happening so fast it’s hard to keep up. I’ve always fancied myself a bit of a futurist, but I could not have predicted the power of building information modeling and how it has transformed the way architects work. The blog posts by Bill Allen and Graham McKay only hint at how BIM and AI will drive the future of the design and construction industries.  

(1)    As Bill Allen points out, AI can competently replicate the works of great artists, such as a Bach cantata or a Prairie-style home by Frank Lloyd Wright, by analyzing representative datasets and utilizing them to generate algorithms that mimic their work.  

No comments: