Real-time Computer Model of the City Center of Ancient Rome

A real-time, interactive computer model of the city center of ancient Rome was created with support from the Andrew W. Mellon Foundation from 2001-2007 as part of the Rome Reborn project. Rome Reborn is an international initiative to create 3D models of the topography, sites, and monuments of ancient Rome illustrating the city's development from the first settlements in the late Bronze Age to its depopulation in the middle of the sixth century CE. Rome Reborn 1.0 was finished in 2007. It shows the city as it might have appeared on June 21, 320 CE, including ca. 7,000 buildings as well as the streets, water features, and terrain within the Aurelian Walls.

o   Project management. The project is managed by the Institute for Advanced Technology in the Humanities at the University of Virginia. Project Director is Bernard Frischer. Associate Director of 3D data capture is Gabriele Guidi (Politecnico di Milano); Associate Director of 3D hand modeling is Dean Abernathy (University of Virginia).

o   Partner institutions have included the INDACO Lab of the Politecnico di Milano, AUSONIUS of the CNRS and University of Bordeaux, the Maison de la Recherche en Sciences Humaines of the University of Caen, and the Experiential Technologies Center at UCLA.

o   Types of evidence and models. The evidence for the topography of ancient Rome can be divided into two categories:

-Class I: these are features that are known in great detail and can thus be modeled with high probability. There are ca. 250 Class I sites and monuments known from late-antique Rome, of which 31 have been modeled to date.

-Class II: these are features whose location, design, and identification can be only approximately known from such evidence as ancient literary and epigraphic sources, the Severan Marble Plan, and the regionary catalogues. There are ca. 6,750 Class II sites and monuments known from late-antique Rome. These have been modeled using a library of architectural forms derived from a study we undertook of the repeated elements in I. Gismondi’s Plastico di Roma antica, a 1:250 scale plaster-of-Paris model of Rome in the Age of Constantine housed in the Museum of Roman Civilization, Rome/EUR.

o   Scientific Advisory Committee. Each Class I feature of the digital model is created with the help of a team of scholarly experts. To date, the following scholars have collaborated: Carla Amici (University of Lecce), Heinz Beste (German Archaeological Institute, Rome), Diane Favro (UCLA), Philippe Fleury (University of Caen), Bernard Frischer (University of Virginia), F. Cairoli Giuliani (University of Rome “La Sapienza”), Lynn Lancaster (Ohio University), Paolo Liverani (Vatican Museums and University of Florence), Martin Schaich (Arctron), Russell Scott (Bryn Mawr College), Robert Vergnieux (CNRS, Bordeaux), Mark Wilson-Jones (University of Bath).

o   Documentation. When exploring the model, it is always possible to open a Documentation window to see the evidence behind the reconstruction. The window provides the following information:

-Class I features: When a Class I feature is finished, the scientific advisors sign a form authorizing its release; the project team fills out a metadata form based on the Dublin Core and a documentation form that records the archaeological evidence behind the reconstruction.

-Class I and Class II features are documented by pertinent articles in S.B. Platner and T. Ashby, A Topographical Dictionary of Ancient Rome (Oxford 1929); the online publication of the Severan Marble Plan; and our digital edition of P. Aicher, Rome Alive: A Source-Guide to the Ancient City (Chicago 2004; corrected version for the online edition, 2008). Bibliography for each Class I and (if available) Class II feature is automatically generated from the OPAC of the German Archaeological Institute.

o   Georeferencing of data. Both the model and its documentation have been georeferenced with ca. 5 meter accuracy. Georeferencing permits the documentation appropriate to a specific location in the model to be automatically displayed when a user so requests.

o   Software. Rome Reborn 1.0 was authored in a variety of 3D modeling software, including 3D Studio Max, Maya, Multigen Creator, and Polyworks. The individual components making up the model were ultimately converted to OpenFlight (.flt) format. The user interface is based on OpenSceneGraph for the model and Google Earth for the documentation. Our Web application platform for the Documentation window includes: PHP 4.4.7 (Hypertext Preprocessor); Apache 2.0 (Server); and KML (Google Map API).

o   Next steps (2008-9):

-Enhancement of the current version of the Rome model through:

×Correction of any flaws and omissions.

×Improving the frame-rate of the model.

×Replacing baked-in textures with shaders.

×Replacement of current Class II models based on a library of architectural forms derived from Gismondi’s Plastico di Roma antica with new, more-detailed procedural models using CityEngine software.

×Inclusion of Class I models supplied by third parties.

-Publication of the Rome model to the Internet through SAVE (“Serving and Archiving Virtual Environments”). SAVE is the first online, peer-reviewed journal where scholars can publish 3D models of cultural heritage sites. SAVE, which grows out of research undertaken at IATH with the support of two grants from the National Science Foundation, will permit 3D models to be accessed by hundreds of users simultaneously logged-in, providing interactive, real-time exploration of models. The models will be made available in different modes and scales and with different functionality, including:

×Global mode will use the GIS-based software package Google Earth to permit the user to see a greatly simplified version of the model positioned over the modern site, which can be toggled on and off; to have access to spatialized documentation available in the other modes.

×Collaborative mode will employ a metaverse software solution (such as Open Sim, Croquet, Wonderland, vel sim.) to put the model into an environment in which users can interact with each other and with software agents such as bots in virtual space. The collaborative mode will permit a moderately simplified version of the model to be: viewed and discussed by collaborative modeling teams dispersed throughout the world during the production phase; peer-reviewed in real time during the editorial phase; used in research and education after it has been published. The user can access global mode and high resolution mode when in collaborative mode.

×High resolution mode (using Mental Images' Reality Server software) will permit hundreds of simultaneously logged-in users to view the model on the Internet at the highest resolution, with real-time rendering of light, with protection of intellectual property, and with the use of tools such as a virtual measuring tape and calipers.

Links

·         Project web page: www.romereborn.virginia.edu

·         New publication by IATH on the use of 3D models as tools of discovery in archaeology: http://www.iath.virginia.edu/~spw4s/Beyond/BeyondIllustration_experiment2_sm.pdf