|
|
Jon Lindsay
Kiersten Muenchinger
Connie Chiueh
John Russell
Tony Vastola
© 1997 Copyright on this material is held by the authors.
At the heart of BIOculars is a single computer simulation of the park's ecosystem that runs on a central server at park headquarters. The simulation includes geographical, climatological, and plant and animal census data. Each time a virtual species is created, the BIOculars device transmits information about it to this central park server.
Virtual animals are active in the park ecosystem, roaming about, eating, perhaps being eaten, reproducing, and dying. Their success or failure is determined by the adaptations given to them by their creators. The resulting virtual ecosystem is thus a complex result of the actions of many BIOculars users and the interactions that arise among their creations. Park visitors simply look through their BIOculars to view the virtual ecosystem. BIOculars transmits its physical location -- utilizing the Global Positioning System (GPS) -- to the central server. The server then sends back information about any virtual animals that might 'live' in the immediate area. If animals are nearby, then the user sees, through BIOculars, stereoscopic 3-D images of virtual animals integrated with images of the natural scene. A user may see, for example, a virtual dinosaur-like creature dart out from behind an actual tree, then hide behind a real rock.
BIOculars is a leap ahead of virtual ecosystems viewed on a computer screen in that park visitors go out into the park to see their creations in action. User testing revealed that families find BIOculars fun and educational in part because it enhances their hiking and exploring activities. If users create virtual gazelles that flourish only in a particular meadow in one end of the park, then the visitors must physically go there to observe the gazelles. Since the virtual ecosystem is a constantly running simulation that is active whether anyone is viewing it or not, things can change greatly from hour to hour. Thus, if visitors create a mouse that thrives on an abundant vine in the park, the population should rise; someone else may later create a species that preys on the mouse, causing its population to plummet. Because BIOculars provides a window onto a virtual ecosystem that coincides with the real one, users learn the effects of changing the balance of nature.
It should be noted that as a concept device, we acknowledge that BIOculars requirements such as size of video display, accuracy of GPS, and computer edge-finding capabilities are beyond current capabilities.
Creating an animal is a core activity of BIOculars. Our latest prototype lets users begin with default animals that can be modified with diverse animal phenotypes characterizing locomotion, skin types, hearing, sight, eating and reproductive habits, etc. The basic animals are located on the right strip, and are selected by spinning the thumbwheel to the desired picture and touching the screen. Modification parts are scrolled through with the left-hand thumbwheel. Touching an animal part causes it to appear on the basic animal. To get information about a specific part, such as its benefits, or its use by real animals, the user simply touches the help button, then the animal part. The user may also use the Random button, to make an animal with randomly selected body parts.
The animal locator, or map mode, shows where a selected virtual species 'lives' on a map of the park.
Species information mode displays details about the population, growth rate, survival outlook, lifestyle, and adaptations of a chosen species. This mode may be triggered on the screen itself, or when BIOculars is in viewer mode by spying an animal, pressing a button, then opening to the screen, which is automatically in information mode.
After visiting several tourist sites, we found parks most compatible with our goals in part because of popular 'interpretive activities' -- campfires, docent walks, and visitors' center exhibits. We thought an interpretive scavenger hunt might work. At that point, we considered our users 'park visitors' rather than 'tourists.'
After more brainstorming, we thought of a system that would allow users to create virtual animals and view them in the real park. Initially, we envisioned this as two separate devices: a creation kiosk located in a visitors' center and a remote viewer. However, role-playing exercises within the design team showed this to be problematic; the viewing activity was too passive, and we wanted to modify and create our animals in the field.
Our 'one device or two' question was answered only by creating several physical prototypes. These included models in which a viewer docked to a portable computer, and models that integrated the viewer and screen. Users found two-device models too 'computerish' and cumbersome. A transforming, binocular-like device for creating and viewing animals was well-received, possibly since binoculars are already familiar tools in the outdoors.
The software interface went through many changes, too, beginning with a rough Director movie prototype that walked users through building and releasing an animal. We tested this interface and a physical model of the device with children in Palo Alto, CA.
The general results for the concept were encouraging, as designing an animal and being able to see it out in the wilderness was popular with the users. (One girl enthusiastically said she would want to make her stuffed animals come to life.) However, the interface contained too much text, and navigation was slow and linear.
We were concerned about how to navigate a large number of menu items for these base animals and their optional parts. When we added physical wheels to scroll through the menu items in a sort of filmstrip on the screen, we discovered it was a way to quickly navigate large menus, and users told us it was fun. It also got away from the graphical user interface convention of pull-down menus, which helped to make our interface less 'computerish.' We then mocked up a new Director interface that used the wheels in other modes, providing consistency across the interface.
For our next test, we chose a semi-natural preserve with a lake, trees, and trails. First, subjects were given a laptop with the BIOculars interface. Then, after creating an animal, they were led on a walk around the lake. They were stopped at locations that we had previously photographed and were given the physical model to look through. Then they were shown the image on the laptop of the scenery before them, but with virtual creatures frolicking about.
On the laptop, scrolling seemed confusing as the thumbwheels were represented on the computer screen. The physical thumbwheels on the model were more understandable, and subjects said they would be much more inclined to play with them. Other problems were caused by limited differentiation between buttons that remained constant and those that changed during each mode. Some users said they wanted to see more realistic animals rather than the cartoonish images shown, but the users were still enthusiastic about the overall concept.
After this test, we revised again. The software interface was cleaned up with sharper, more defined images, and the global navigation buttons were differentiated from the local buttons. Eyepiece ergonomics were changed to fit small and large hands.
|
|