click to view contact details click to search using Google SiteSearch click to view directions to the lab
Sichuan Virtual Dig's papers
Sichuan Virtual Dig main page

where we are
contact us
research projects
news & events
hitlab nz
hitlab au




Sichuan Virtual Dig

This project has been completed. While the area of research may continue to be of interest, the HITLab is not engaged in a current project in this area. This page may contain dated information.

The Seattle Art Museum and the University of Washington Human Interface Technology Laboratory recently completed a year-long collaboration to create a virtual archeological dig. The Virtual Dig ran from May 10, 2001, to August 12, 2001 as part of the Sichuan China artifact exhibit. During that time more than 25,000 people experienced this novel interactive experience.

The Virtual Dig combined HI-SPACE and ARToolKit interaction technologies along with 10 networked computers, 6 cameras, and 6 projectors. This page focuses on the interactions supported by the technologies. Each interaction is labeled as HI-SPACE or ARToolKit to identify which technology supported the specific interaction. Additional content on the Virtual Dig can also be found here.

The Virtual Dig was designed to be an interactive learning experience to help museum goers understand where the artifacts exhibited in the galleries came from. In the experience a narrator guides attendees through a series of explorative interactions. The story line moves forwarded based on participants completing simple tasks. Time outs are also used if needed to keep the experience moving.

Click on any of the images or text below to play a short QuickTime clip showing each of the Virtual Dig segments.

The basic process of the Virtual Dig is one of uncovering and examining artifacts found at an excavation site. The sections below take you through the main interactions of the Virtual Dig in the same order attendees at the Seattle Art Museum experienced them. For the Virtual Dig, real images of an actual dig sites were used. All models were created by 3D scanning replicas of actual artifacts. The models were then textured with images of the artifacts.

The Room <1.86MB>

The Room

The Virtual Dig consists of three interaction stations. The stations have two large screen displays, the tabletop itself and the wall behind the table. Each station provides the same 10-minute experience for up to 10 participants each. Attendees of the Virtual Dig do not have to wear any of the traditional VR equipment such as head-mounted displays or trackers. This allowed for a continuous flow of over 120 people per hour.

A Magical Start <200KB>

A Magical Start: HI-SPACE

A museum guide would give an introduction and then start the interactive experience. We wanted an innovative way to start the experience but one that would protect against accidental starts. The method we settled on had the guide place an object of a certain size on the table. Once the object is detected, the system looks for a sequence of hand gestures. These gestures trigger the start of the experience. This not only provides a safe method to start the experience but also amazed the museum goers with its magic show like qualities.

Brushing <2.13MB>

Brushing: HI-SPACE

As the first task in the Virtual Dig, participants are asked by the narrator to help in excavating a new site in the Sichuan province. The tool they are instructed to use is a brush. This is an ordinary 1-inch paintbrush. Because the video based tracking system uses normal objects and not specialized devices, almost anything can be used to interact with the computer. Participants use the brushes to remove grass and dirt to expose the first artifact layer.

Examining Tusks <1.44MB>

Examining Tusks: ARToolKit

The first artifact layer is composed primarily of elephant tusks. The reason for the tusk layer is an open question for archeologists and the narrator expounds on this fact. Participants are instructed to pick up their other tool, a shovel. Each shovel has a unique pattern that is recognized and tracked by the system. As users move their shovels over the table (excavation pit) an tracked image of each shovel is shown on the wall display. On the image of each shovel is a tusk. There are different tusks, some broken, burnt, or otherwise damaged. These represent the condition the tusks were in when originally found by the archaeologists. By rotating and tilting the shovel the corresponding tusk image is also rotated and tilted allowing for examination.

Removing Tusks <876KB>

Removing Tusks: HI-SPACE

It is now time to move to the next artifact layer. The participants are instructed to again pick up their brushes. Different groups of tusks are highlighted and can be removed by brushing over them. As each successive group of tusks is removed, more and more of the next layer is exposed. There are six tusk groups that need to be brushed away.

Examining Masks <1.15MB>

Examining Masks: ARToolKit

The second and final layer in the pit contains many artifacts. Among these are several types of masks. Just like the tusks, the masks have been damaged in different ways. The Participants are again instructed to use the shovels to pick up and examine the masks just as they did with the tusks. There are five different masks modeled. They show not only different masks styles but again different types of damage as well. The narrator discusses some of the different types of damage that can be seen on the masks. Because the number of participants can exceed the number of masks modeled, duplicates will appear at times.

Finding Statue <1.2MB>

Finding Statue: HI-SPACE

One of the most impressive pieces of the Sichuan exhibit is an eight-foot tall bronze statue. Originally the statue was found in pieces in the pit and later repaired for display. The two main pieces of the statue are visible in the pit. Recognizing the statue pieces can be difficult. To help find the statue pieces participants are instructed to use the brush. Underneath the brush will appear a semi-transparent spot. The spot is blue when a brush is not near the statue pieces and turns red as the brush moves closer to the statue pieces. This hot-cold game helps participants find the statue pieces. Once two or more participants have found a statue piece all pieces were highlighted and the spots disappeared.

Combining Statue <980KB>

Combining Statue: ARToolKit

Using the shovels, the participants are directed to pick up the two pieces of the statue and put them together. As before, the statue pieces appear on the shovels in the wall display. This time only two shovels are used and bringing the two shovels close together combines the pieces.

Fitting Objects <1.1MB>

Fitting Objects: ARToolKit

The hands on the statue are unusual and appear to be designed to hold something. Exactly what that something is remains an archeological debate. Speculation is that a tusk seems to be the best fit to the curved hands. In this final interactive sequence, participants are given the chance to place different objects found it the pit into the statues hands. Placing a shovel over the objects picks them up. Placing the shovel holding the object next to the shovel holding the statue will place the object in the hand of the statue. Since it is believed the tusk is the correct object, all incorrect objects will not quite fit. The narration describes why the objects do not correctly fit in the hands of the statue.

Sponsoring Agencies

The Seattle Art Museum


Peter Oppenheimer <peter at>, Mark Billinghurst <mark.billinghurst at>, Richard May <rmay at>