Academic Research in Architectural Computing

Back in the days when I was doing my Ph.D. in Architecture at UC Berkeley, I sometimes met practicing architects who were surprised that there even was something like a Ph.D. in Architecture and that someone would actually bother doing it. But architecture is as much of a social and cultural phenomenon and a science as it is a professional activity, and in many universities across the world, there have been long-established research programs that study different aspects of architecture such as history, social and cultural factors, building science, material science, and design theory. With the advent of computers in the 1970's, a new area of research opened up for architectural academics: how computational technology can better assist the design process. Since then, many innovative ideas in architectural computing have been explored by academic researchers including generative design systems; precedent-based design tools; evaluation and analysis tools for all aspects of building design, including "soft" criteria such as form, habitability, circulation, etc.; tools for capturing design intent; and so on.

Even the concept of a building data model, as in BIM, is not new. It has, in fact, been the Holy Grail of architectural computing research right from the mid-'70s, when the earliest attempts were made to develop integrated design systems that could support a suite of applications capable of operating together rather than just individually. A single building model was the central piece around which such integrated systems—dealing with all aspects of design, construction, and facilities management—could be built, and several models were developed for prototype systems by different research teams. It is interesting to see some of these ideas finally make it to commercial tools used by practitioners.

Many more interesting ideas are still being explored in architectural computing research at universities around the world, and at least some of them will be translated into commercial tools in the not-too-distant future. It should be of interest to practitioners to know more about what these ideas are, so that they can be better prepared for future technological developments in the field and embrace them more readily. In this "Building the Future" article, AECbytes explores the architectural computing research being pursued at four leading universities around the world. It is the first in a periodic series of articles that will look at computing research not only in architecture, but also in engineering and in construction.

Georgia Tech, USA

Prof. Chuck Eastman of Georgia Tech is one of the academics with the most visibility in the industry, and under his guidance, a wide variety of research activities are being pursued in the College of Architecture at Georgia Tech, with many of them making it to commercial development. A prominent example of this is the CIS/2 standard, which is widely deployed in the steel fabrication industry (see the recent AECbytes article, The CIS/2 Format: Another AEC Interoperability Standard). Georgia Tech has been contracted by the AISC (American Institute of Steel Construction) to support the deployment of CIS/2 as well as provide technical support, testing, tutorials, and point upgrades. Georgia Tech is also working on another related effort, the harmonization of CIS/2 with IFC, the other leading interoperable product model in the AEC industry (see Figure 1). The work involves identifying "use cases" where there is a clear need for IFC and CIS/2 to interoperate, providing extensions to both IFC and CIS/2 so they can interoperate for the specified use cases, and develop a translator that supports exchange for the specified use cases. (For more on the IFC, see the AECbytes feature, The IFC Building Model: A Look Under the Hood.)

Another significant project that Georgia Tech has been involved in is the development of a new precast concrete 3D parametric modeling system, in collaboration with Tekla, a Finland-based CAD company, and the Precast Concrete Software Consortium (PCSC). Georgia Tech contributed to this project in many ways, including development of the process modeling technology that allows automatic capture of information flows and derivation of the supporting product model, developing the diagramming methods for domain experts to depict parametric model design behavior, coordinating the development of industry-wide parametric modeling specifications as well as specifications for the top-down multi-functional parametric model design tool, and defining the detailed product specification and review of final product release. The first commercial version of this product was released in August, and the first precast projects designed with it are now being constructed (see Figure 2).

Other projects include developing methods to reduce the development time of product models from the current 5-10 years to 1.5 years or less; relational database implementation of product models like IFC and CIS/2, the Intellectual Property rights for which were sold for exclusive use in the steel fabrication industry in 2004; developing the import and export modules for exchanging structural analysis data between a 3D modeling system and structural analysis packages using IFC; developing new user interface guidelines of complex 3D CAD systems for a CAD company; and developing the practices and advancing the tools for Product Lifecycle Management (PLM) in the construction industry. A recent initiative at Georgia Tech, in response to the industry's transition to BIM (building information modeling), is the AEC Integration Laboratory, which will serve as a center for activities such as assessing emerging workflows, testing systems for scalability and robustness, and preparing a new generation of construction IT technologists. One of the additional projects being pursued under the framework of this laboratory is effective design for fabrication (DFM), an example of which is shown in Figure 3.

And finally, Georgia Tech also has a strong interest in computational building simulation, with a dedicated research group involved in projects such as understanding building ventilation and using CFD (Computational Fluid Dynamics) tools effectively for improving health and quality of indoor environments, developing a Building Performance Assessment Toolkit, risk analysis of indoor air quality and mold growth in buildings, and an automated space layout tool that meets the optimum design requirements and maximizes design quality in terms of design preferences.

Eindhoven University of Technology, Netherlands

The Eindhoven University of Technology is one of the leading institutes pursuing architectural computing research in Europe. One of its long-standing research interests is Design & Decision Support Systems (DDSS) for Architectural Design and Urban Planning. Within this broad area of emphasis, it is pursuing several specific research projects, some of which are described below.

Multi Agent Systems for Cooperative Design: "Multi Agent Systems" are modular, independent and autonomous software programs as opposed to traditional, "monolithic" software development, and are especially promising for distributed work environments with loose hierarchical structures. This research project explores the potential of Multi Agent Systems for use in architectural design and the building process as a whole, assisting designers to easily get information related to a particular design task and to automatically offer solutions to design problems. Based on information that is harvested from different local and remote resources, new strategies for solving design problems are proposed to the user by individual agents.

User Simulation of Space Utilization: The aim of this project is to develop a method to simulate how space is utilized in order to provide reliable data on human movement for performance evaluation (see Figure 4), instead of the assumptions that are currently made about people's movement through space and their responses to the environment. These assumptions are input for important design decisions (e.g., capacity of elevators, width of corridors, escape routing, etc.) and sophisticated calculations (e.g., lighting simulation, airflow simulation, evacuation simulation, etc.).

Multi Agent Model for Network Decision Analysis: This project aims to develop a multi-agent model for simulating pedestrian activity and movement patterns. Movement is simulated using a grid and steering behavior. Variability is introduced by superimposing agents, with their specific agenda, environmental knowledge, beliefs, and choice heuristics and scripts. The results of the simulation are visualized in a virtual reality environment.

Computational Representations of Words and Associations in Architectural Design: This project explores design support through the use of words, in particular by means of offering verbal associations based on the annotations written by the architect. The hypothesis is that this will lead to increased creativity of the design and reduced fixation in design process, since architects use many kinds of representations in the early design phase such as sketches, marks, images, and annotations. Annotations, in particular, provide additional information about the design that can not be captured by the other representations.

The Eindhoven University of Technology was also closely involved in the development of the Building Management Simulation Centre, a new training centre in the Netherlands for construction management (see Figure 5). Unlike traditional construction management training, it does not take place in practice but in a high-tech environment in order to have maximum control over the training circumstances and to collect as much data as possible about the trainee, during the training. The Centre consists of a virtual building site, a (real) site hut for the trainees, and a control room. A new course system was also designed which intermediates between the trainee, the trainer and the building site. A central role in the course system is played by so-called transitions, which are composed of activities that have to be performed by the trainee. An activity is fulfilled by entering web-based forms with the right content and in the right order. Interaction with the trainee is also established through actors that interrupt the planned flow of activities. After a training session, the system can reproduce all actions that were taken; it can show the consequences for the building and present the construction management performance. The Centre is being used by several architectural and civil engineering schools.

RMIT University, Melbourne, Australia

Architectural computing research at RMIT is pursued in the Spatial Information Architecture Laboratory (SIAL), which is a broad multi-disciplinary center that seeks to integrate technological developments in areas such as multimedia, sound scape systems, mechatronics, computer graphics and interfaces, materials technology, manufacturing processes, collaborative virtual environments, animation, computer games, virtual reality facilities, GIS, and online environments. It is associated with all areas of design at RMIT, combined with social and cultural studies.

Not surprisingly, a wide variety of research projects in diverse areas are pursued at SIAL. Some key projects from an architectural/construction standpoint are Flexible 3D Modeling, which uses CATIA to create flexible (associative geometric) computer models for design and links these to physical modeling and prototyping; Atmosphere, which explores the use of game engines as a dynamic design media supporting new forms of representation; Eureka, which explores nonlinear systems for presentations, encouraging the presenter to navigate freely through multimedia assets allowing fluid contextual narratives to emerge; and Re-engineering, which explores cross-disciplinary design collaboration between architects and engineers using high level parametric design techniques. The research pursued at SAIL has been applied to many real-world projects, including the Shoal Fly By series of public art for the Melbourne Docklands precinct; a new footbridge to provide access to the development of the St Helier waterfront area on the Island of Jersey; and the Australian War Memorial in London (see Figure 6).

Other notable SIAL projects include a combined design studio with MIT School of Architecture and Design and Gehry Partners; a multimedia-based design decision support environment for students, researchers and practitioners; investigations into the ontology and representation of design; and interpreting Gaudí's drawing for the Passion Façade design for construction during the coming years. The laboratory also acts as a creative think-tank accessible to both local and international practices, including Arup in Melbourne and London, dECOi in Paris, and Gehry Partners in Los Angeles.

University of California at Berkeley, USA

The focus of computer-aided architectural design research at UC Berkeley is on two related topics: studying the theory and practice of creative, multidisciplinary collaboration and developing digital means that can support and enhance it; and exploring the impact of digital technologies and telecommunication on the practice and products of architecture.

The interest in collaboration arises from the reality of architectural design as a social activity, in which the expertise of many professionals is needed to complete even the most trivial building. The research focuses on understanding how these professionals collaborate, and finding computational means to assist them. This includes the development of shared product models that can serve as means of high-level communication among the participating professionals, and the development of collaborative design environments where the participants can come together virtually, as needed, even when they are separated spatially and temporally. My own Ph.D. research (described briefly in the following section) was an example of a shared product model, while other examples include the Cube Game, an application that allows different professionals to collaborate in the design of single family houses, and virtual places, the digital equivalent of physical venues that can support a multitude of human activities where collaboration among design professionals can be explored.

Virtual place-making is also being used to explore the affordances of cyberspace for collaborative social and professional action, based on principles drawn from architecture, film, philosophy, social sciences, and more. Two areas have been chosen to focus the research efforts: education and cultural heritage. The choice of these areas was partly informed by the needs of each one and the opportunities offered by cyberspace, and partly by the availability of grants and collaborations with others on and off campus. According, virtual learning environments have been developed to demonstrate that the practice of place-less learning, also known as "remote education"—which separates students from their peers and from the context of the educational process and transforms the learning experience into an exchange of the learned content alone—can be improved by incorporating architecture's place-making principles. In particular, it would be of value when the content of the learned material was consistent with the context of its learning, such as in the study of historically significant environments—ancient cities and villages that no longer exist and, therefore, cannot be physically visited. Accordingly, models were developed of the city of Cairo in the 11th century (see Figure 7), of a Neolithic village in Turkey that existed 9,000 years ago, and more recently of 7th Street in Oakland, California, as it existed in the 1950s, when it was the Jazz capital of the Bay Area and the cultural hub of its African American society.

This interest and research focus has lead UC Berkeley, over the past 3 years, to establish a new inter-disciplinary center, called the Center for New Media (CNM), for exploring the implications of technology on many disciplines who, under the typical administrative structure of academic departments and colleges, are too isolated from each other to be able to consider the issue in a cross-disciplinary manner. The Center includes about 90 faculty affiliates from 31 departments on campus, serves some 250 students each year, engages in multi-million dollar research activities, and organizes 6-8 events per year. The explicit intent of the Center is to understand the effect of computing technology (broadly defined) on culture and society, and to pass on that understanding to students who must learn to operate within this fast evolving reality. To support such knowledge acquisition and transfer, a new Designated Emphasis and Graduate Group in New Media has been established, where Ph.D. students from every department on the Berkeley campus now have the opportunity to study and pursue research in this area.

Conclusions

My own Ph.D. research work at UC Berkeley was focused on developing a data model that represents, in an integrated fashion, both the spatial and structural components of the building so the same model can be used to plug in architectural-analysis tools, structural-analysis tools, and tools that can evaluate the synthesis of space and structure. It was aimed primarily at the schematic design phase, where the most crucial design decisions are made and the need for design support is the greatest. I implemented a working prototype of this model, comprising a simple 2D graphical editor for drawing both the spatial and structural components of the building, and query functions that read the input file generated by the editor and automatically retrieve the geometrical and topological information about the building from it that would be needed by tools evaluating different aspects of the building design. An apartment design evaluator supported by the system was also developed as proof of concept. The link to existing commercial evaluation tools was demonstrated by automatically deriving a building geometry input file from the system for DOE2, a sophisticated energy-analysis application.

While I am delighted to see some of the ideas that I explored in my research in today's BIM tools, some other capabilities in my prototype solution are still not available commercially. Similarly, there are many other relevant ideas out there in the research community that have still to see the light of day. This calls for a better synergy between university research and industry implementation in the building industry and the technology industry that supports it. In other fields of engineering and computer science, both technology and its implementation are very advanced compared to technology for AEC primarily because of the close ties between universities and the industry. We need to replicate this in the AEC industry. Academics doing architectural computing research should not remain content to develop their theories and concepts with no regard to implementation; at the same time, the industry needs to be more aware of and sponsor research work in architectural computing at universities and implement those concepts and ideas into tools for the profession at large.

Acknowledgments

My sincere thanks to all these researchers and faculty members for their help with this article: Prof. Chuck Eastman and Prof. Godfried Augenbrow of Georgia Tech; Prof. B. de Vries of the Eindhoven University of Technology; Andrew Maher and Chi Sung Foo of RMIT; and Prof. Yehuda Kalay of UC Berkeley.

About the Author

Lachmi Khemlani is founder and editor of AECbytes. She has a Ph.D. in Architecture from UC Berkeley, specializing in intelligent building modeling, and consults and writes on AEC technology. She can be reached at lachmi@aecbytes.com.