A 3D Common Ground: Bringing Humanities Data Together Inside Online Game Engines

July 17, 2013, 08:30 | Short Paper, Embassy Regents E

Over the past decade humanities scholars have begun to seriously explore the potential of using 3D modeling software to reconstruct historic spaces and architecture. However, because of technological and design limitations, their works have tended either to feature visually rich environments that were closed and static, or more interactive and data driven spatial projects that failed to seize upon the potential of realistic 3D imaging. But now as 3D gaming engines are making it easier to display visually sophisticated virtual environments in browser windows in real-time, and to access and interface with large data sets, new possibilities exist for 3D reconstructions. Using both demos of historical reconstructions running live in the Unity 3D engine and mockups of future plug-ins, this presentation imagines future 3D historical reconstructions that bring a variety of data together, and argues for projects that are persistent, historical, collaborative, and curated.

The use of online 3D engines for the display of historical reconstructions first allows for persistent environments that can constantly be explored by any number of users. These reconstructions might be a single town, building, or other reconstructed site that would exist perpetually on the web, constantly accessible to the public. Previously projects aiming for a high level of detail and realism would have to store their 3D scenes offline, and could only share their work by rendering image stills or animation clips that would show prerecorded views. However, significant improvements in the graphical quality of 3D engine browser plug-ins mean that now, these scenes can be uploaded to the web, and users can walk and fly though historical spaces freely, and even interact with the environment.

The sophistication of these 3D engines also opens up new possibilities for depicting historical change over time. The customization of game engines means that scene elements like individual buildings can easily be assigned time ranges, allowing for an environment that can show growth and changes in the architecture and landscape as users move backwards and forwards through virtual time. These innovations are made possible because 3D gaming engines handle complex, visually realistic 3D models more efficiently than platforms like Google Earth or ArcGIS, meaning scholars no longer must choose between a realistic, but static reconstruction, or a historical GIS scene that showed change over time but could not support advanced imaging.

Because browser engines are more adept at dynamically displaying change within a 3D scene the possibilities for scholarly and community collaboration have enlarged considerably. First, with only a little additional scripting, 3D engines can be customized to access large outside data sets, like those used with Google Maps API or standalone GIS programs, but they also can allow any visitor to annotate virtual space in real-time. In a persistent online reconstruction of historic church for instance, users could walk down the aisles and through the balconies, while pausing to mark spots, leaving virtual footnotes in 3D space, which could then link to outside data, text, or images. Additional scripting could store these user annotations in a database, and allow for the community accumulation of data from different scholars and the public, within the same persistent environment. With more advanced plug-ins, there is even the future possibility of users manipulating the environment itself from inside a browser, adding or modifying architectural elements as part of their own interpretations.

Because these persistent, historical, and collaborative 3D reconstructions could quickly become cluttered with multiple interpretations, dedicated editors who could curate the virtual space could provide a balance between open contributions and consistent quality. First, individual contributions or annotations could be grouped into different layers, which viewers could choose to show or hide, and certain of these layers could be ranked both by user voting, as well as by editor endorsement. Especially in the case of contributions that involved significant changes to the 3D models in the scene, site editors could even act as or coordinate with peer reviewers, allowing for scholars to gain publication credit for adding especially valuable 3D content or other data.

This presentation features as an example, a real time reconstruction of an 18th century North American imperial fort running in a customized version of the Unity 3D game engine. This reconstruction shows the development of fort’s architecture over time through various stages of construction, disrepair and remodeling, and also features multiple interpretations of particular architectural elements based on conflicting pieces of evidence. Within this evolving 3D environment, the presentation also show sets of annotations in 3D space that link to copies of archival documents, maps, plans, and archeological artifacts. In addition to plotting historical events or the locations of artifacts, these points may act as footnotes linking to particular pieces of evidence that informed the look or design of a given part of the interpretation. A custom feature programmed into the 3D engine could also allow individual users to tour the space and plot their own original points, each of which can link to multimedia content. Finally, this demonstration will explore possibilities for the importation of Google Earth points, GIS data, and topographic maps into the example 3D reconstruction as distinct viewable layers.