State-of-the-art of tactile Hyperglobes Andreas RIEDL, Email: [email protected] University of Vienna, Department of Geography and Regional Research Universitaetsstrasse 7, A-1010 Wien, Austria

Abstract This article focuses on the successor of virtual globes namely tactile hyperglobes. These tactile hyperglobes allow – as opposed to the two dimensional representation of Geo-Browsers (e.g. Google Earth, Virtual Earth ,…) – the earth’s presentation in her natural and three-dimensional appearance. This makes it easier for the user to link the information on the globe to the reality that the globe is representing. The Department of Geography and Regional Research of the University of Vienna invested in the beginning of 2005 in a tactile hyperglobe. Therefore this department is the first European research facility which focused research activities on the visualization of global topics under the use of spherical displays. On the one hand this paper will give insight in advantages and disadvantages of different types of tactile hyperglobes from a technical perspective and on the other hand introduces the functionality of OmniSuite, software dedicated to tactile hyperglobes.

1 The terminology of digital globes The following definition applies in the same way to both traditional analog globes and digital globes: "A globe is a scale-bound, structured model of a celestial body (respectively firmament) presented in its undistorted three-dimensional wholeness" [RIE-00]. In general globes can be distinguished regarding their implementation by three parameters [RIE06]: • The nature of the cartographic image (analog, digital), • The character of the globe body (physical, real) and • The kind of representation space (real, virtual). Besides the well-known analog physical globe the following categories of digital globes can be identified [RIE-00]: • Virtual hyperglobes: Visualization of the digital image on a virtual globe body in virtual space. • Tactile hyperglobes (material hyperglobes): Visualization of the digital image on a physical (touch-sensitive) globe body in real space. • Hologlobes: Visualization of the digital image on a virtual globe body in real space.

Figure 1: Categories of digital globes. (slightly modified, Riedl 2006) A three-dimensional model is a basic requirement for digital globes. But the emphasis is on “its undistorted three-dimensional wholeness”. Therefore, today’s overused term “digital globe” for applications like Google Earth or Virtual Earth 3D is valid only to a certain extent. More precisely, these “Geo-Browsers” serve as a kind of interface to large-scale geo-data or maps. There is a stronger emphasis on detailed information of earth’s surface (or other celestial bodies) and less on the globe as a whole with its global interdependencies. Geo-Browsers produce maps in the broadest sense of the term, mostly in some perspective view. Therefore a geo-browser is a geo-multimedia application, where a digital globe (virtual hyperglobe) is one of his features. For in depth discussions on the term of (digital) globe refer to Hruby et al. (2009), Riedl (2000, 2008), Scheidl (2009).

Figure 2: Tactile Hyperglobe (Technorama, Winterthur, Switzerland - installed by Globoccess 2008).

2 Variants of tactile Hyperlobes The era of tactile hyperglobes started in 1992 with the “GeoSphere Globe” in Brazils’ space research center. This was the first globe (2m in diameter), which offered the possibility to project global images onto a translucent sphere. Contrary to today’s projection systems the projectors were not made for permanent operation. Therefore a translucent satellite image was superimposed on the spheres surface in order to show an image when the projection system was turned off. Subsequently different technical approaches have been tested concerning tactile hyperglobes. Today one can speak of small-scale industrial production. In general tactile hyperglobes have in common that the display serves as globe-body as well as visual display unit. The following techniques for projecting the image onto the globe can be differentiated (Riedl 2008): Outside-projection: Systems with an outside-projection use a perspective azimuthal projection for mapping the image on the sphere. Typically the system consists of four beamers positioned along the sphere’s equator. Sometimes a fifth or sixth projector is used for mapping the pole regions.

Figure 3: Outside-projection system (top-view). Inside-projection: Systems based on an inside-projection use a special azimuthal projection for projecting a world map through a hole at the sphere’s bottom via optical system onto an acrylic glass ball. The optical system is based either on special lenses (fisheye) or on a convex mirror. Fisheye-based hyperglobes are limited to a single beamer, whereas mirror-based systems are able to use two beamers simultaneously (doubling the resolution).

Figure 4: Inside-projection systems, fisheye-based (left) and mirror based (right). Direct-projection: We speak of direct-projection when the spherical display is both, projector and screen. They are currently not available. Flexible OLED-displays may pave the way, but there are still five to ten years to go. Spherical displays with direct-projection would have the best image quality: no pixel distortion, no blind spots on the globe, no shadowing of the projection beam, high resolution. OutsideProjection Resolution Image Quality Installation/Deinstallation Ambient light insensitiveness Space requirment Hardware requirement Blind Spot Shadowing of projection beam

+++ +++ ++ +++ -

Inside ProjectionFisheye + + ++ + +++ + ++ +++

Inside Projection Mirror ++ ++ + + +++ + + +++

Tab. 1: Pros (+) and cons (-) of projection techniques (modified Riedl, 2008) Furthermore the globe-body can be either “solid” or “inflatable”. Whereas inflatable displays are primarily intended for mobile applications and solid displays for permanent installations. Actually there are about 10 companies, which offer spherical displays resp. tactile hyperglobes.

3 Authoring and presenting global Stories The biggest advantage of a globe in general is that it is distortion free and shows spatial relationships found in the real world. There simply does not exist another cartographic product that comes as close to a globe from this perspective. Contrary to virtual hyperglobes, in the case of tactile hyperglobes one is in front of a real (scale-downed) 3d-model of the earth (or any other celestial body). Seeing a tactile hyperglobe in action is fascinating. Like being an astronaut and looking from outer space on earth, or in other words: “Oh, my God! Look at that picture over there! Here’s the Earth coming up. Wow, is that pretty!" (Frank Borman, commander Apollo 8) In order to create gripping content for tactile hyperglobes some tasks have to be done. On the one

hand, data needs to be projected onto a 3D-display geometrically correct. On the other hand, the results of this projecting process have to be controllable and accessible in a user-friendly way. The projection problem may be considered as a cartographic basic task: The base maps for global topics are typically in an equirectangular projection (plate carrée). This global map gets translated into a special azimuthal projection (fig. 5). This azimuthal projection includes the path of rays and recurring reflections inside the globe. Here ends the routine, because the rendering must be done in real-time to give functional-associative correct feedback to the users in reference to their manipulations via direct manipulation interface.

Figure 5: Projecting process for a tactile hyperglobe (single beamer) The Hyperglobe-Research-Group at the Department of Geography and Regional Research (University of Vienna, Austria) is developing an Authoring and Presentation Software dedicated to tactile Hyperglobes called OmniSuite. We make use of C++ and the open source ObjectOriented Graphics Rendering Engine (OGRE) that offers numerous relevant features, e.g. the combination of different layers or the support of different graphics file formats (Hruby et. al. 2008). For in depth information on programming and software engineering related to OmniSuite see Kristen (forthcoming 2009). OmniSuite’s 3D-engine is responsible for projecting and rendering the content in real-time according to the user interactions and used globe-system (see chapter 2). One focal point of the software was the development of a (globe-)platform independent 3D-engine. Therefore a set of needed parameters had to be identified and implemented in different settings, some of the adjustable parameters are: • Display-settings (touchscreen, globe-projector(s), external info-screen) • Initial-image: default story, sweet spot

• • •

Projection settings (fig 6): scaling, offset, horizontal adjustment … Image-quality/performance tuning: mesh resolution, cube map resolution, antialiasing … Debug info: performance status, log files

Figure 6: Projection settings Another focal point of our research group is creating content for the globe. OmniSuite’s StoryEditor (authoring module) provides tools for content creation according to a storyboard. The content has to be specific to the users needs. Therefore we take in consideration feedback from visitors of our tactile hyperglobe (1,5m in diameter). As a result the presentation of themes is usually made in a dynamic way via animations and in combination with short stories like explanations. Those “stories” makes it easier for users to understand causal interdependencies and processes. The story library of experienceable themes includes amongst others: • Effects of earth-sun relationship and earth rotation (seasons, course of the day, timezones) • Surface of the earth (topography, satellite imagery) • Geology themes like continental thrift (from 600 Mio years in the past to 100 Mio in the future) • Climate and oceanography (real time weather, hurricanes, snow line in the course of the year) • Ecology (climate change, El Niño, air pollution) • Traffic (shipping traffic, air traffic) • Economy (fishing areas, crop yield, economic power) • Alliances, pacts, military power • Extraterrestrial themes (sun, planets, moons, constellations) • Historical globes

• •

Advertisement, promotion, public relations Digital art and entertainment

Figure 7: OmniSuite’s controller interface (master-mode, dark-skin design) Furthermore OmniSuite gives the story author the possibility to implement interactive functionalities that allows manipulation by the user regarding the three interaction categories (orientation and navigation, altering the visual appearance, information retrieval). These include: • Assigning an image or image sequences (e.g. global maps) to a certain layer (defaultlayer, legend-layer …) • Specifying a layers extension (default-layer across the whole globe, legend on a section) • Defining an animation (duration, speed, loop, forward/backward) • Referencing audio tracks • Switching layers on/off • Defining bookmarks • Controlling the globe’s rotation Typically user interactions with the globe are performed via graphical user interface (GUI) on a touchscreen. Implementing a GUI can be done in many different ways. The simplest one is by using OmniSuite’s interface templates and varying it with skins (fig. 7). More experienced authors can alter the templates or design complete new GUIs based on html/css or flash.

Conclusion The field of applications of tactile hyperglobes ranges from representation purposes, public relation activities, installations in museums, digital art exhibitions up to edutainment and didactics. One of the intriguing installations of a tactile hyperglobe can be seen at the Technorama in Winterthur (Science Center, Switzerland). With tactile hyperglobes users are able to “see and feel” the globe in reality like an astronaut. Hence a close emotional relationship with matchless quality of experience evolves in a natural way. This matchless quality of experience is proved by users feedback. Therefore after 500 years of globe history (Erdapfel, M. Behaim, 1492, oldest preserved globe),

respectively 2200 years (Krates von Mallos, 200 B.C., only written evidence) globes will have a new prosperous era.

References Beck A. (2006), Google Earth and World Wind: remote sensing for the masses. In: Antiquity 80, 308 – http://antiquity.ac.uk/ProjGall/beck/ Blower J., Gemmell A., Haines K., Kirsch P., Cunningham N., Fleming A., Lowry R. (2007), Sharing and visualizing environmental data using Virtual Globes. In: Proc. of the UK e-Science Meeting 2007, S. 102–109. Hruby Florian, Kristen Jürgen, Riedl Andreas (2008), Global Stories on Tactile Hyperglobes – visualizing Global Change Research for Global Change Actors. In: Proceedings, Digital Earth Summit on Geoinformatics: Tools for Global Change Research, Potsdam, Germany, 2008. Hruby Florian, Miranda Ruth, Riedl Andreas: (2009), Bad Globes & Better Globes – multilingual categorization of cartographic concepts exemplified by “map” and “globe” in English, German and Spanish. Proceedings, 24. ICA Cartographic Conference, 2009. Santiago, Chile, 2009. Kretschmer I. (2000), Globenforschung in Österreich. In: Riedl A. (2000), S. 141–155. Kristen J. (2009), 3D Grafikprogrammierung interaktiver kartographischer EchtzeitAnwendungen – am Beispiel eines taktilen Hyperglobus. Wien, Univ. Wien, Dipl.Arb.(forthcomming). Möller M. (2008), Geobrowser – Katalysatoren für Geoinformation im Unterricht. In: Jekel T., Koller A., Donert K. (Hrsg.), Learning with Geoinformation III – Lernen mit Geoinformation III, S. 159–170. Heidelberg. Riedl A. (1999), Virtual Globes. A New Era for Globes? In: Proc., 19. ICA Cartogr. Conf., Ottawa, 1999, S. 649–656. Riedl A. (2000), Virtuelle Globen in der Geovisualisierung. Untersuchungen zum Einsatz von Multimediatechniken in der Geokommunikation (mit einem Nachwort „Globenforschung in Österreich“ v. Kretschmer I.). (= Wiener Schriften z. Geogr. u. Kartogr., 13). 158 S. Riedl A. (2005), MUGL - Multimediales Globenmuseum. In: Mitteilungen der Österreichischen Geographischen Gesellschaft, Bd. 147. Wien, 2005. S. 265 - 282. Riedl A. (2006), Digital Globes. In: Cartwright W., Peterson M.P., Gartner G. (Hrsg.), Multimedia Cartography, Second Ed., S. 255–266. Berlin – Heidelberg, Springer Verlag. Riedl A. (2008), Entwicklung und Perspektiven von Taktilen Hypergloben. In: Mitteilungen der Österreichischen Geographischen Gesellschaft, Bd. 150. Wien, 2008, S. 340-356 Riedl A. (2009), Taktile Hypergloben – die nächste Stufe in der Globenevolution. In: Kriz K., Kainz W., Riedl A. (Hrsg.): Geokommunikation im Umfeld der Geographie. Wien, Institut für Geographie der Universität Wien, 2009 (=Wiener Schriften zur Geographie und Kartographie, Band 19)

Scheidl W. (2009), Virtuelle Globen im Unterricht – Multiperspektivische Untersuchung des didaktischen Potenzials von Google Earth und NASA World Wind im Geographieunterricht Allgemeinbildender Höherer Schulen, Master Thesis, Donau Universität Krems, 127 S. Schleicher Y. (2008) Geobrowser-Didaktik – ein Beitrag zur aktiven Erweiterung des topographischen Wissens. In: Jekel T., Koller A., Donert K. (Hrsg.), Learning with Geoinformation III – Lernen mit Geoinformation III, Wichmann, Heidelberg, S. 60–69 Schratt A., Riedl A. (2005) The Potential of Three-Dimensional Display Technologies for the Visualization of Geo-Virtual Environments. In: Proceedings, 22. ICA Cartographic Conference, 2005. A Coruña, Spain