NSF funds research to put powerful virtual environment technology into the hands of architects and engineers
Blacksburg, Va. -- Virtual environments computer-generated worlds that create the sensation of being in a place, whether an insect's gut or a room of one's own design -- seem like a logical tool for architectural design or building construction. Immersive virtual environments (VEs) use spatial interaction rather than requiring the user to interact with a design using a mouse and keyboard, for example. But building designers rarely uses VEs.
VE software tools have not been particularly intuitive or user friendly, and do not include high-level functionality, says Doug Bowman of Blacksburg, assistant professor of computer science in the College of Engineering at Virginia Tech. He is addressing the problem by designing three-dimensional (3D) interaction techniques and a software framework explicitly targeted to design and construction.
His work and teaching has now been enhanced by a $500,000 NSF CAREER grant to implement a unique software framework he calls iDesign.
As part of the CAREER project, he is developing immersive design applications with the iDesign software in structural engineering, interior design, above-ceiling coordination, and architectural design.
Bowman first became aware of the disconnect between designers and the possibilities offered by immersive virtual environments when he was a graduate student at Georgia Tech "We used head-mounted displays -- two tiny screens in front of your eyes," he explains. "It feels immersive because the head tracker measures head movement so that the virtual world moves when your head moves."
He worked on conceptual design of space to create a 3D sketchpad for architects. "It was interesting to us the VE group in computer science -- but no architect would have used it because we didn't know how to do 3D user interfaces. For example, on your desktop you have menus and toolbars that let you interact. In 3D, we had nothing like that."
His next project was better. "We worked with Zoo Atlanta on a project to teach environmental design students how to design an animal habitat. We provided them with a virtual environment based on an existing habitat, then created tools that allowed them to move trees, add rocks, change the terrain by selecting from eight models, and change the public viewing areas. They could do a complex design in a short amount of time and it made sense."
But there are still problems with generic 3D interaction techniques, Bowman says, which gave him the idea for the CAREER proposal -- the application of domain-specific interaction techniques to take advantage of the properties of a particular domain.
For example, if you are an interior designer and want to place furniture in a virtual room, with existing object placement techniques, you can select an object from a list or some distant place and set its 3D position and orientation. But you don't want a chair to float above the floor, be upside down, or penetrate a wall. "You don't need to be that expressive," says Bowman, "So we add constraints. The chair must always sit on the floor and only rotate on one axis. We are exploring interaction techniques that are designed specifically for the needs of design and construction applications."
In addition to interior design projects, Bowman is working with Walid Thabet, associate professor of building construction in the College of Architecture and Urban Studies (CAUS), who wants a tool to design mechanical, electrical, and plumbing systems within buildings, ranging from homes to large office buildings.
He has also been working with Mehdi Setareh, associate professor of architecture at Virginia Tech, who is interested in how building structures respond to environmental effects, such as earthquakes. (See Shaking Things Up In Virtual Reality, http://www.research.vt.edu/resmag/2002summer/virtual.html).
An initial project for Bowman's NSF CAREER funded research in early 2003 was a task called cloning. "Mehdi wants to build structures that are large and complex, requiring an enormous number of elements beams, columns, slabs. Even with efficient object creation techniques, building a frame piece by piece is going to take more time than you want to spend wearing VE headgear. So, we allow the designer to create a small piece or common unit that is going to be repeated throughout the structure, then clone, or replicate it as needed in 3D space."
To fulfill the educational component of the NSF CAREER award, Bowman is the lead author on the book, 3D User Interfaces: Theory and Practice (Addison Wesley, 2004), which will be used in a new advanced topics computer science graduate course on 3D interaction. He also teaches a graduate course on virtual environments in which students perform semester projects in VE interface design.
Source: Eurekalert & othersLast reviewed: By John M. Grohol, Psy.D. on 21 Feb 2009
Published on PsychCentral.com. All rights reserved.
It is common sense to take a method and try it; if it fails, admit it frankly and try another. But above all, try something.
-- Franklin D. Roosevelt