Listen to an AI-generated, text-to-speech version of this article

In part one of our exploration into computational thinking (CT) within the architecture, engineering, and construction (AEC) industry, we discussed the foundational principles of CT and how it seamlessly bridges human creativity with machine precision. We highlighted its transformative nature, especially in bolstering efficiency, fostering innovation, and enhancing collaboration between humans and technology.

As we transition to part two, we’ll expand on this groundwork. We’ll delve deeper into CT’s profound impact on engineering practices, how it amplifies an organization’s collective intelligence, and showcase real-world applications — including some ground-breaking initiatives right here at Henderson. Join us as we continue our journey illustrating how CT is reshaping the future of the AEC landscape.

Engineering

Various industries and fields of study have developed specialized methods for acquiring knowledge, such as the scientific method used for conducting research in the various branches of science. While CT can be applied in many fields, it has been particularly instrumental in engineering, providing a structured approach to aligning the level of detail in equations with the complexities of real-world challenges.

In the AEC industry, the problems engineers face can be visualized on a spectrum of complexity. At one end of the spectrum, we find the atomic simplicity of abstraction: at the other, a plunge into infinite detail. Historically, engineers have predominantly operated within a narrow band in the middle of this spectrum. It is in this band that CT plays a crucial role, not only optimizing tasks within this specific zone but also providing tools to navigate the uncertainties encountered when exploring either extreme of the spectrum.

Despite the public perception of the AEC industry as a breeding ground for unconventional and avant-garde architecture, the reality is that it is steeped in conservatism—a stance justified by the inherent risks and liabilities involved in construction. This conservative approach means that many construction projects lean towards employing proven concepts, materials, and techniques. Consequently, the bulk of engineering work in the AEC industry is firmly anchored in the mid-range of the complexity scale, a domain where the tried-and-tested prevails, and innovation is approached with caution.

For such situations, there are a wealth of established and well-documented workflows, tested extensively across varied scenarios. This accumulated expertise, when meticulously recorded, serves as a naturally balanced dataset prime for developing automation tools. The CT process often begins with data collection, so in these situations the task is simplified, allowing for intensified focus on subsequent stages. Sometimes, the established workflows can also be repurposed as the framework for an automated tool. While directly converting human procedures into computer code might overlook potential efficiencies, at a minimum it can serve as a guiding foundation for the logic design process.

For engineering tasks approaching the extreme ends of the complexity spectrum, the methodology of CT remains unchanged, but the benefits undergo subtle transformations. Regardless of a challenge’s position on the scale, CT’s robust toolkit is adept at managing any arising ambiguity or convolution. Armed with analytical tools like decomposition, pattern recognition, and abstraction, even the most unprecedented or anomalous challenges become significantly more manageable.

Collective Intelligence

CT grows an organization’s collective intelligence by fine-tuning both its capacity and capability to undertake a myriad of tasks and address diverse problems. The foundational step involves utilizing CT to establish a unified language for problem-solving. Once established, it facilitates the development of tools that streamline all three phases of the intelligence cycle: the acquisition, processing, and utilization of information.

The acquisition of information, which represents the first phase of the intelligence cycle, often entails labor-intensive and repetitive efforts that demand sustained accuracy, making it inherently stressful. To alleviate this, basic automations can be deployed for routine tasks such as data validation, cleansing, and migration. Advanced tools like natural language processing and computer vision offer solutions for more complex tasks. For instance, optical character recognition can be employed to assist with data extraction.

During the second phase of processing information, the implementation of various artificial intelligence data analysis tools can significantly enhance efficiency. Machine learning (ML) algorithms excel at tasks like data clustering and classification, offering not only efficiency but also the ability to uncover non-linear or non-obvious relationships among data variables that might elude human analysts.

The final phase, utilizing information, presents a substantial opportunity for optimization through CT. Traditional manual techniques fall short of exploring the full potential of advanced design methods such as generative and parametric design. In this context, applying ML algorithms allows for rapid iteration through multiple design options, pinpointing the most optimal solutions efficiently.

By optimizing each phase of the intelligence cycle, CT fosters a richer collective intelligence within an organization, enabling it to navigate the complexities of tasks and challenges with increased agility and insight.

Example Applications

CT has manifested as a transformative approach in the AEC industry, enhancing design, construction, and building management processes. Numerous practical applications have already demonstrated its success and versatility. A prominent area of application is in calculations and simulations designed to predict or analyze building performance. Techniques such as building energy modelling, electrical short circuit analysis, and building load calculations all employ CT to construct mathematical and computational models simulating system behavior.

CT is also being used to create AI-driven customized facilities management solutions. A typical example is the integration of machine learning for personalized lighting and window treatment controls, contributing to enhanced energy efficiency and occupant comfort. AI also optimizes resource allocation and sustainability through predictive adjustments, enhancing resource utilization.

One application that is still in its infancy is evolutionary optimization using client-values-based design logic. This technique calibrates ML algorithms during the design process based on client preferences, evaluating factors like cost, time, environmental conservation, and human comfort. Together with building codes and local legal constraints, these client values delineate boundary conditions, guiding the identification of optimal solutions.

A celebrated instance of CT application in the AEC industry is the design of the Morpheus Hotel in Macau by Zaha Hadid Architects. This 40-story building is distinguished by an external structural exoskeleton which obviates the need for internal walls or columns. Remarkably, the design employed the Grasshopper visual programming language for parametric modelling, which proved to be twenty times faster and reduced material usage threefold compared to traditional methods.

Morpheus Hotel in Macau

CT also underpins the development of commercially available design tools like Autodesk’s project refinery. This tool extends Autodesk’s generative design capabilities, allowing architects and engineers to explore diverse design options and meet client requirements by setting and optimizing goals. Another noteworthy tool is Hypar, an interactive generative design platform for automated building systems design. It enables the quick creation of conceptual-level designs based on user-defined logic and features an API for exporting selected design concepts directly to Revit.

Here at Henderson, we harness CT to develop innovative design tools such as the duct automation tool. This tool leverages generative design to create modular duct systems, automating the HVAC system design process and seamlessly integrating design for manufacturing and assembly (DfMA) as a standard design procedure. We also recently utilized CT to develop the office space automator, which automates the design office spaces in Revit for the electrical, A/V, fire protection, mechanical, and telecom disciplines.

Final Thoughts

By now I hope you will agree that computational thinking stands out as a transformative catalyst, driving innovation and reshaping the landscape of the AEC industry. By combining human ingenuity with computational precision, CT is enhancing collective intelligence within organizations, and paving the way for a multitude of applications, ranging from pioneering architectural designs to the development of user-centric tools.

The possibilities of how CT can be applied are vast and varied, which speaks to its versatility and potential to deliver sustainable and tailored solutions. As the industry continues to evolve and technology advances, CT is set to play an increasingly significant role in the design, construction, and management of buildings, marking a computational revolution. It’s not a prerequisite for every engineer to “learn to code,” but proficiency in CT is becoming essential to remain competitive in this dynamically evolving landscape. The trajectory we are on is clear: CT is not merely a technique but a philosophy, opening new avenues and redefining the boundaries of what is possible in the AEC industry.

Considering this, as industry leaders and innovators, we bear a responsibility to fully embrace and exploit the myriad potentials of CT. We must actively engage with this transformative paradigm, push computational frontiers, and collaboratively sculpt a future where the convergence of technology and creativity yields a built environment that is increasingly sustainable, harmonious, and responsive to the diverse needs of our evolving world.

More from Henderson on Computational Thinking

Part 1: Computational Thinking in the AEC Industry