News Release

Structural Engineers and Computer Scientists Hope to Integrate Disciplines to 'Revolutionize Building Construction'

San Diego, CA, October 02, 2008 --   Despite the breathtaking evolution of computer science over the past 30 years, structural engineering — hindered by a reluctance to adapt to digital innovations — has remained relatively unchanged as a discipline. With this in mind, a group of UC San Diego engineers and computer scientists, together with their counterparts at Germany's Bauhaus University of Weimar, are hoping to completely reform the way buildings are constructed by integrating the two fields and bringing data visualization and online collaboration to the fore of structural design.

Jurgen Schulze Oct 08
Calit2 project scientist Jurgen Schulze navigates through a 3-D virtual rendition of the San Francisco/Oakland Bay Bridge in the institute's StarCAVE.

More than 30 representatives from both universities met at the UCSD division of the California Institute for Telecommunications and Information Technology (Calit2) last week for a "Collaborative on Technology and Society" that focused on Integrative Structural Design, or the merging of sensing, modeling, simulation and visualization into one cohesive design paradigm. Organized by Jacobs School of Engineering Dean Frieder Seible, the workshop featured more than 15 presentations by experts in both fields, including Associate Professor of Computer Science and Engineering Ingolf Krueger, Associate Professor of Structural Engineering and workshop chair Falko Kuester, Calit2 project scientist Jürgen Schulze and Maurizio Seracini, the director of Calit2's Center of Interdisciplinary Science for Art, Architecture and Archaeology. Also in attendance were Bauhaus University of Weimar (BUW) Vice Chancellor Karl Beucke and Marcel Urban, director of the Engineering Sciences Division for Deutsche Forschungsgemeinschaft, the German equivalent of the National Science Foundation.

"Right now, we have very non-interactive ways of constructing buildings," explained Urban. "But computer visualization is becoming the hot spot in structural engineering. It starts from the architectural plans and carries through to construction and building to doing diagnostics on existing buildings. If we can begin from a macro-scale view of a structure and zoom in to the nanoscale, it might revolutionize building construction."

Frieder Seible
Frieder Sieble, Dean of the Jacobs School of Engineering and a structural engineering professor, is hoping the collaboration will allow for multi-scale structural engineering, which will allow engineers to examine structure from the macroscale to the nanoscale.

Added Seible: "Although there's a widening gap between what is being done in research and what is being done in practice, there's an increasing notion of generating computer graphics and automatically integrating them into structural analysis. If we can put these pieces together, we can come up with a completely different approach to how we analyze structures."

Putting that approach into concrete terms was one of the goals of the workshop, which outlined, presentation by presentation, the challenges inherent to the field and the solutions that computer science might provide. Carsten Koenke of Germany's Institute of Structural Mechanics relayed his wish for robust algorithms that can convert structural data (geometric models) into image data (numerical models), and would allow engineers to examine structures from a macro- or meso-scale all the way to a micro- or even nanoscale.

Professor Frank Werner, chair of the Institute for Steel Design and Construction at BUW, pointed out that most structural design models are based on simple empirical formulas that account for certain influencing parameters, such as direction of stress, working temperature, and quality of weld.

 "But at the moment," he noted, "it's impossible to connect all of these parameters into one computer model." Instead, he said most engineers have to conduct three separate simulations — process, material and structure simulations — to determine a building element's load-bearing capacity. Each simulation can take a week or more, which underscores the need for a quick, digitized method for simultaneously analyzing all facets of a structure.

Such a method might be developed via the imaging and visualization capabilities at Calit2 and UCSD's Jacobs School of Engineering. In his presentation, Seracini discussed some of the tools and techniques CISA3 is already using to analyze the structural history and viability of aging buildings in Florence, Italy, including thermography, ground penetrating radar, neutron activation analysis, neutron radiography and neutron tomography — this in addition to more traditional imaging techniques, such as x-ray, gamma ray, near-infrared and acoustic imaging.

"These technologies would allow for multi-scale engineering, where every crack and every rebar would become visible," Seracini said. "This, I think, would be a dream come true."

As for visualizing and displaying those images, Schulze and Kuester discussed the potential for using Calit2's immersive, interactive 3-D StarCAVE and tiled HIPerSpace wall (the largest in the world) to view structural models on large, super-high-resolution displays. Scientists at the institute have already put their techniques into practice with an application that displays computer-aided design models of parts of the new San Francisco-Oakland Bay Bridge. The application allows users to walk/fly through actual-size elements of the bridge to find material clashes, construction errors, and generally to draw conclusions about whether the structure could be built as designed. Also serving as an example was the design model of the Jacobs School's Structural Engineering / NanoEngineering / Media Arts building, for which the groundbreaking ceremony is less than a month away.

By the end of the workshop, which featured two days of presentations, facilities tours and collaborative working sessions, the group had identified three research arenas key to solving the challenges laid out by the presenters: 1) Developing mathematical foundations of models and visualization; 2) creating integration and interactive models and 3) developing a means to assess existing structures.

Beucke pinpointed the first arena as key to integrating structural engineering with advances in visualization, noting that "the mathematical foundations are very important."

"We need to agree on a common foundation," he added, "and I think it's vital that we use visualization technologies to do this. Researchers should be excited by these developments, and not see them as a burden. I have to say, I'm a little frustrated that structural engineering is still very traditional. The textbooks haven't changed in 30 years. I would like to see technology-driven new solutions."



Falko Kuester Oct 08
Calit2 project scientist Jurgen Schulze navigates through a 3-D virtual rendition of the San Francisco/Oakland Bay Bridge in the institute's StarCAVE.

The problem with developing those solutions is that engineers and computer scientists "don't often speak the same language," said Calit2 Professor of Visualization and Virtual Reality Falko Kuester.

"There is much data collected from computer scientists, but not so much from structural engineers," he added. "They're completely different in terms of data storage and data mining, and historically, we have lived in disjointed worlds. If we can come up with models that allow us to tie both of these domains together, we can take a significant step forward."



Integral to this collaboration will be cyberinfrastructure, which will allow the scientists and engineers to communicate and share information using efficient, high-tech applications that are not limited by geographic location. Providing examples of currently functioning online collaborative models were Krueger, Sylvia Mazzoni of the University of California, Berkeley, and Lelli Van Den Einde of the San Diego Supercomputer Center and the Jacobs School's structural engineering department.

Krueger suggested that structural engineers consider developing service-oriented cyber-architecture, which he described as a "paradigm for organizing and utilizing distributed system capabilities that may be under the control of different ownership domains." Mazzoni's lecture on OpenSees (Open System for Earthquake Engineering System) encouraged structural engineers to develop a similarly powerful, flexible and open-source tool for the simulation of structural systems. Likewise, Van Den Einde's talk demonstrated the benefits of NEESit, which develops software and online services to promote earthquake engineering research and communication.

Added Van Den Einde: "The challenge is that engineers are used to doing research in their own ways, and they're reluctant to digitize and share their data. But with digitized information, we can start developing visualization capabilities. We need to be able to tie together computational simulations plotted with experimental data in a 3-D view. We're not there yet, but there are some people in the field who are very much ready for this."

"Research is becoming more and more global," Beucke added. "I'd like to see something like NEESit made available to developing countries. How can we bring them into the picture? It benefits them to be able to collaborate, but we would also benefit. If they are doing better, we are doing better."

The Technology & Society Workshop was the second in a series designed to expose engineers and scientists to various engineering topics and demonstrate how the field contributes not just to research, but to society on the whole. Visualization was the topic at last year's workshop, held at BUW. That meeting paved the way for the discussion this year, and the participants expect that the Collaborative on Technology and Society between UC San Diego and BUW will continue with a meeting in Weimar next year.

"The specific topic for next year's workshop is still to be defined, but will largely depend on the actual collaborations that develop between researchers in the two institutions as a direct result of this year's workshop at Calit2," Dean Seible stated.

"The workshop this year has been interesting both from a scientific point of view and a collaborative point of view," Urban remarked. "Ideally, I'd like to see the creation of an international research training group or a small PhD course in integrative structural engineering."

Added Kuester: "This workshop has been a full success in bringing an interdisciplinary team of researchers, practitioners, and students together to chart the path towards Integrative Structural Design, evaluating the first critical building blocks and forming a team committed to moving this forward. Being both a computer scientist and structural engineer, witnessing the creation of this type of enabling synergy is very exciting."

Media Contacts

Tiffany Fox
Qualcomm Institute