STANDARDS PIPELINE
Vol.32 No.4 November 1998
ACM SIGGRAPH
Imaging and Graphics Business Team Addresses Opportunities for Standardization
George S.Carson
GSC Associates, Inc.
November 98 Columns
George S.Carson
The preceding Standards Pipeline (Computer Graphics 32(3) August 1998) introduced you to the Imaging and Graphics Business Team (IGBT) formed by Joint Technical Committee 1 of ISO, the International Organization for Standardization, and IEC, the International Electrotechnical Commission (ISO/IEC JTC 1). The purpose of the business team is to identify opportunities for creating new international standards by working with business, government, public sector organizations, graphics and imaging experts, as well as developers of standards outside of ISO and IEC. For more information, please feel free to consult the team’s Web site.
The IGBT has formed four working groups to address opportunities for standardization in the following areas:
- New interaction
- Synthetic environments
- Information archiving and distribution
- Fundamental objects
This month’s Standards Pipeline describes the first two of these areas of potential standardization. The first, interaction, is described in a contribution written by Dr. Paul ten Hagen of CWI which was submitted to the IGBT as a starting point for the New Interaction Working Group.
Discussion of the second area addresses the need for standard representations of natural or artificial environments, particularly from those in the simulation and modeling, entertainment and instruction/training fields. Such representations’ components are located in reference to a well-defined spatial origin and coordinate system.
Towards an IT Standard for Interaction
Dr. Paul ten Hagen
CWI
Introduction
Current advances in presentation technology have brought a wealth of nicely visualized information both into homes and onto desktops. This achievement is the result of the well-orchestrated application of new techniques for data representation, integrated display processing, data compression and networking. Much of this orchestration was achieved through open, high-level object-oriented programming technology. While we have seen the first generation of multimedia platforms reach homes and offices, we have also observed a lack of effective interaction tools required for end users to adequately deal with the flood of information from the new media. Because of this, growth in the market has slowed, creating lower than expected returns for the content creation industry.
The relative success of presentation technology owes much to the emergence of a commonly accepted architecture that has helped blend and focus innovation. This architecture is partly due to the pioneering work of standardization initiatives from both industry and academia. The natural question today is: “Can we extend such an architecture for interaction?” This short essay aims to address this question, explaining how such an effort might start, and how it relates to current achievements in presentation technology standards.
Interaction Devices
It is worthwhile to briefly consider the concepts of input and interaction. The discussion throughout this paper will be at a conceptual level. No technical details will be elaborated upon at this stage. A device such as a keyboard can be used both as an input device (by typing in new text) and as an interaction device (by moving the cursor through key strokes to the target string for editing, for example). The difference between the two is the role of feedback. Feedback for an input device merely echoes the keystrokes on the screen. The cursor in one way or another highlights the selected text string amidst the others; hence, we interact with data already presented.
The concept of interaction invalidates the classical concept of input and output as two independent streams. An interaction device is fundamentally characterized by the control it exercises over the relation between input and output. An interaction device is not sufficiently described as a special kind of input device. It is just as well described as an output device. Thus, interaction devices are hybrid devices combining both functions.
The functionality that determines controlled relations in interaction needs to be thoroughly understood. This is the area where standardization is deemed necessary. In the following we will highlight aspects of this functionality in order to introduce the areas for study. Following this we will be in a better position to describe the relevance of such a standard both for (new) application areas as well as for platform independent interaction.
Finally we will assess the difficulties anticipated when creating a standard for interaction functionality. Highly relevant to this discussion is the possibility of achieving interoperability with existing standards.
Input-Output Relations
The classical input process transforms the raw input actions performed by an end user or by an autonomous input device, such as a camera, into meaningful entities for the receiving program. A typical example of an input process is a recognition process. Similarly, a classical output process transforms internal abstract data into perceivable presentations. This process refines higher level entities into lower level atomic presentation actions. A typical example of an output process is a rendering process.
We will illustrate the concept of input-output relations with examples from possible relations between recognition and rendering. A recognition process is hierarchical in nature, for example: lexical, syntactic and semantic interpretations; image contouring; or texture classification, segmentation and interpretation. Similarly, the rendering process can be viewed as a cascade of decompositions and transformations towards primitive collections of elements that can be fed into the presentation hardware. Even the presentation hardware pipeline itself has such characteristics.
International Standard 14478, Presentation Environment for Multimedia Objects (PREMO), includes a reference model for the output hierarchy and to some extent for the input hierarchy for multimedia. It also provides a framework for compound devices for multimodal input/output (I/O). One could say that PREMO defines the internal relations between compound input devices and compound output devices, as do some other specifications. Hence, using such a model one can synchronize several output modalities, such as picture and sound.
State of the art interaction requires compound hybrid devices. Each subdevice in the compound device must be capable of controlling a portion of the I/O streams. Moreover it must be able to accept externally provided data which may influence the way it behaves. For instance, a speech recognition system which uses a dictionary to enhance reliability must be able to use locally provided dictionaries. A real-time 3D scene renderer must accept view directions from an input device controlling the viewpoint. The movements of a vehicle in a 3D simulation might be controlled by spoken commands. The vehicle to be controlled in this manner might be the one that was selected by focussing the view on it. If I do not have a voice system available, it should be replaced by another option (e.g. dials) without having to change the underlying program. Perhaps in this case the speed of the vehicle would automatically adapt to a less efficient control technique.
In order to enable user friendly and efficient interaction, the optimal context must be generated. The mechanism for context realization is also based on controlling the interaction device behavior through impulses from the context. Whether this is done implicitly, e.g. by caching techniques, or explicitly, by making the context a visible part of the presentation, should be the decision of the interaction designer. If they want to reuse an interaction device in a different context, they should be able to do so without reengineering. A special case of this occurs when higher-level interaction devices determine the context of the lower level subdevices. Similarly, the other way around it should be possible to adapt the context to the interaction to short or long term history. Thus context impulses are bidirectional.
It is possible to extend current reference models such as PREMO in this direction. One goal of further work should be to clarify to what extent this is worth doing in a standardized fashion. It should be noted that this way of looking at interaction devices would widen the horizon for new devices as well as new combinations, e.g. graphics and image processing.
New Application Classes
A simplified characterization of the external behavior of a classical interactive application is to present the results of an ongoing calculation and to accept user input to influence further calculations. If the reaction of the system is real time, we call this computational steering. Steering is the current state of the art in I/O intensive interactions. It has revealed the power and potential of giving the user the sensation of being immersed in the application environment. Current applications are mostly concentrated at the high end of computing, such as large-scale engineering applications incorporating simulations. The resulting immersion provides a more efficient transfer of information between the user and the system. This is the reason why simpler forms of immersion should be brought to the desktop and home PC.
Some examples of application classes where interactivity can be improved are:
- Navigation through and updating of large information bases, where part of the interaction resources are used to maintain the user’s filter of interest. Such information systems require built-in links to presentation and navigation services that can accommodate a wide variety of platforms. A specific example is a Web browser that can connect with any local interaction device(s).
- Person to person communications, where all participants dwell in the same information domain.
- Distributed interactive information systems, where the resulting interaction is composed from the interaction facilities of each subsystem and is transparent to the user.
Many more examples can be given, but the above serve as good indicators of the massive scale on which improved interactivity can be applied.
New Features
Although some of the new features anticipated have been suggested earlier, here is a longer (yet certainly non-exhaustive) list of those deserving further study:
- Dynamic context: Intensive interaction can be streamlined with appropriate contextual help. However the stronger the context, the better the explanatory features in the context must be.
- Behavior rather than predefined dialogue: Dialogue scenarios become dynamic and non-deterministic because the interpretative rules are distributed over independent agents each of whom may trigger a reaction independently.
- High degree multimodality: Combining media both on the input and output sides, and merging them into hybrid devices.
- Recognition process as a part of the input mechanism: This is possibly one of the most significant technological improvements that must be enabled and become economically viable. The raw technology is available, but current software and hardware platforms are not yet ready for it.
- Automatic tracking and recording: Intensive interaction sessions create many reasons for tracking and recording. These include reuse, taking a second look at some significant phase, and remembering points of interest. Implementation implies that daemon processes accompany interactive sessions, making themselves visible on demand.
- On the fly assistance: This can be thought of as an on-line spelling checker for complex interaction entities.
All of these features can be realized through forms of external and internal input/output relations.
Why a Standard?
There are several reasons why a standard is desirable:
- Users will benefit from a competitive market where they can choose their platform from the offerings of competing suppliers and still be able to run highly interactive applications without excessive investments in custom interaction devices or in learning new techniques.
- A user should be able to connect the interaction device repertoire of a given platform to the “interactors” of plug-in applications. Moreover, several such interactors should be able to be combined to create the interaction setting for the platform.
- Efficiency requires that interaction combine seamlessly with compression techniques. This may put another burden on the shoulders of compression experts.
- A user should be able to personalize his or her interaction options and thus create optimal personal ways of working.
- For our discipline to progress, we need to consistently move from the level of abstract composite values to the level of processes that stand for dynamic values, i.e. to a new level of abstractions.
Dr. Paul ten Hagen
CWI
Amsterdam, The Netherlands
Difficulties to be Expected
The basis for this future standards work must be an agreement on a taxonomy of interaction, leading to a common class of basic abstract objects and methods. Currently, all individual application examples are ad hoc implementations. After we have such a taxonomy, the next step is for wide acceptance to be obtained. We should propose the reference model first, and then proceed with the next steps, such as a RFT or NP.
Some of the difficulties that must be overcome are:
- Unfortunately, the output first approach that has resulted from recent advances in presentation technology has created many entangled complex output processes that are insufficiently prepared for interaction. For instance some of these processes cannot easily produce the intermediate results required for relation to an input entity.
- Operating system and windowing support must become more open to real time and resource requirements. For instance repaint policies must become more consistent across systems.
- We need to find generic interfaces to semantics or contents that can filter out relevant data, e.g. a search engine embedded within an interaction handler.
- Strong, intensive interaction tends to break up the application in more fragments. A way must be found to add structure and consistency to such fragments.
- The role of cognition in interaction must be better understood so that it can be stated in terms of design principles.
- The process of developing an interaction standard along these lines must be made attractive enough to industry, such that a sufficient number of companies will see enough profit potential to justify participation in the effort.
- Tightly coupled I/O as foreseen here is even on the conceptual level understood only by a small number of experts.
Conclusion
This discussion has raised more questions than it has answered. My last question: Will we start working on interaction now or shall we leave it to another century?
Synthetic Environments
George S. Carson
GSC Associates
Scope
A synthetic environment is a representation of a natural or artificial environment, whose components are located in reference to a well-defined spatial origin and coordinate system. For representation of the natural environment, this origin is generally either the sun or earth, however any celestial body (or even imaginary body) can be used. An important special case is that of geo-spatial locations referenced to the earth (including its surface, oceans, atmosphere and near space.) A synthetic environment includes terrain and terrain features (both natural and man-made); models of objects such as avatars and certain localized features of the environment with complexity in structure or behavior (e.g., vehicles, buildings, smoke plumes and tornadoes); the ocean (both on and below the surface); the ocean bottom including features (both natural and man-made) on the ocean floor; the atmosphere including environmental phenomena; and near space. In addition, a synthetic environment includes the attributes of the objects in the environment, as well as the relationships among the types of objects. The representation includes constraints necessary to ensure correct automatically generated behaviors. These include topological and rule-based constraints.
The purpose of the Synthetic Environments Working Group of the IGBT is to address all aspects of support for the creation, authoring and interchange of synthetic environments. The scope of the working group includes, but is not necessarily limited to, support for the creation of applications in these areas:
- Simulation and modeling, including simulation based design
- Entertainment, including single user and multi-user games
- Instruction and training
The title “synthetic environments” is chosen to be neutral to:
- Media (versus, for example, “graphical environments”)
- Technology (for example, “world” is a VRML specific term for a similar concept)
It also avoids the over-used term “virtual” (for example, “virtual environments”).
For these purposes, modeling includes the descriptions of the objects within such environments, for example their graphical and/or aural manifestations. It also includes selected aspects of the behavior of the objects within the environment, including interactions among the simulated entities, as well as interactions between the environment and such entities, where the entities respond to events in the synthetic environment or influence the synthetic environment. The mechanism(s) to be used for behavior description are to be determined, and will likely include at least scripting and programming language based representations.
Specific aspects to be addressed or considered include:
- Representational polymorphism
- Conceptual models of the represented objects
- Data interchange formats
- APIs for reading, writing and/or accessing synthetic environments
- Real-time interchange of environmental information
Today, synthetic environments are created through a costly and time-consuming authoring process resulting in a platform-dependent database that supports a single application. One goal of the standards to be proposed by this working group is to enable reuse and sharing of such data between authoring systems, thereby eliminating the need to recreate each database from scratch and also enabling the creation of a market for lower-cost, shared synthetic environments and their components.
George S.(Steve) Carson is President of GSC Associates of Las Cruces, NM, a systems engineering consulting firm specializing in real-time signal and information processing systems. He is the Chairman of ISO/IEC JTC I/SC 24 (Computer Graphics and Image Processing) and has been involved in ANSI and ISO standards development for 20 years.
George S.Carson
GSC Associates, Inc.
5272 Redman Road
Las Cruces, NM 88011
Tel: +1-505-521-7399
The copyright of articles and images printed remains with the author unless otherwise indicated.
Goals and Deliverables
- Determine if it is feasible to adopt international standards in the area of synthetic environments. This includes determining:
- The maturity of technical approaches to synthetic environments
- The degree of international consensus on the best technical approach(es)
- The benefits (or detriments) of standardization in this area
- As with all IGBT work, the plan is wherever possible to:
- Identify and harvest work already proven in commercial practice outside ISO and IEC
- Transpose that work into international standards making only necessary and mutually agreed changes
- The working group will determine the feasibility of using the proven work on the SEDRIS (Synthetic Environment Data Representation & Interchange Specification) project as a basis for one class of standards (non-real-time interchange among authoring systems) to be proposed by this working group. For more information about SEDRIS, you may visit the SEDRIS web site.
- Identify all other feasible base documents and approaches (if any exist).
- Write one or more New Work Item Proposals (NPs) for standards development to support synthetic environments.
The work of the Synthetic Environments Working Group will be completed in 1998. If possible, a New Work Item Proposal will be forwarded to JTC 1 for ballot. The leader of the Synthetic Environments Working Group is Karen Williams of the U.S. Government’s National Imagery and Mapping Agency. You can contact her for more information: kwilliam@MSIS.dmso.mil, or visit the working group’s www site.
Imaging and Graphics Business Team Addresses Opportunities for Standardization
George S.Carson
GSC Associates, Inc.
George S.Carson
The preceding Standards Pipeline (Computer Graphics 32(3) August 1998) introduced you to the Imaging and Graphics Business Team (IGBT) formed by Joint Technical Committee 1 of ISO, the International Organization for Standardization, and IEC, the International Electrotechnical Commission (ISO/IEC JTC 1). The purpose of the business team is to identify opportunities for creating new international standards by working with business, government, public sector organizations, graphics and imaging experts, as well as developers of standards outside of ISO and IEC. For more information, please feel free to consult the team’s Web site.
The IGBT has formed four working groups to address opportunities for standardization in the following areas:
- New interaction
- Synthetic environments
- Information archiving and distribution
- Fundamental objects
This month’s Standards Pipeline describes the first two of these areas of potential standardization. The first, interaction, is described in a contribution written by Dr. Paul ten Hagen of CWI which was submitted to the IGBT as a starting point for the New Interaction Working Group.
Discussion of the second area addresses the need for standard representations of natural or artificial environments, particularly from those in the simulation and modeling, entertainment and instruction/training fields. Such representations’ components are located in reference to a well-defined spatial origin and coordinate system.
Towards an IT Standard for Interaction
Dr. Paul ten Hagen
CWI
Introduction
Current advances in presentation technology have brought a wealth of nicely visualized information both into homes and onto desktops. This achievement is the result of the well-orchestrated application of new techniques for data representation, integrated display processing, data compression and networking. Much of this orchestration was achieved through open, high-level object-oriented programming technology. While we have seen the first generation of multimedia platforms reach homes and offices, we have also observed a lack of effective interaction tools required for end users to adequately deal with the flood of information from the new media. Because of this, growth in the market has slowed, creating lower than expected returns for the content creation industry.
The relative success of presentation technology owes much to the emergence of a commonly accepted architecture that has helped blend and focus innovation. This architecture is partly due to the pioneering work of standardization initiatives from both industry and academia. The natural question today is: “Can we extend such an architecture for interaction?” This short essay aims to address this question, explaining how such an effort might start, and how it relates to current achievements in presentation technology standards.
Interaction Devices
It is worthwhile to briefly consider the concepts of input and interaction. The discussion throughout this paper will be at a conceptual level. No technical details will be elaborated upon at this stage. A device such as a keyboard can be used both as an input device (by typing in new text) and as an interaction device (by moving the cursor through key strokes to the target string for editing, for example). The difference between the two is the role of feedback. Feedback for an input device merely echoes the keystrokes on the screen. The cursor in one way or another highlights the selected text string amidst the others; hence, we interact with data already presented.
The concept of interaction invalidates the classical concept of input and output as two independent streams. An interaction device is fundamentally characterized by the control it exercises over the relation between input and output. An interaction device is not sufficiently described as a special kind of input device. It is just as well described as an output device. Thus, interaction devices are hybrid devices combining both functions.
The functionality that determines controlled relations in interaction needs to be thoroughly understood. This is the area where standardization is deemed necessary. In the following we will highlight aspects of this functionality in order to introduce the areas for study. Following this we will be in a better position to describe the relevance of such a standard both for (new) application areas as well as for platform independent interaction.
Finally we will assess the difficulties anticipated when creating a standard for interaction functionality. Highly relevant to this discussion is the possibility of achieving interoperability with existing standards.
Input-Output Relations
The classical input process transforms the raw input actions performed by an end user or by an autonomous input device, such as a camera, into meaningful entities for the receiving program. A typical example of an input process is a recognition process. Similarly, a classical output process transforms internal abstract data into perceivable presentations. This process refines higher level entities into lower level atomic presentation actions. A typical example of an output process is a rendering process.
We will illustrate the concept of input-output relations with examples from possible relations between recognition and rendering. A recognition process is hierarchical in nature, for example: lexical, syntactic and semantic interpretations; image contouring; or texture classification, segmentation and interpretation. Similarly, the rendering process can be viewed as a cascade of decompositions and transformations towards primitive collections of elements that can be fed into the presentation hardware. Even the presentation hardware pipeline itself has such characteristics.
International Standard 14478, Presentation Environment for Multimedia Objects (PREMO), includes a reference model for the output hierarchy and to some extent for the input hierarchy for multimedia. It also provides a framework for compound devices for multimodal input/output (I/O). One could say that PREMO defines the internal relations between compound input devices and compound output devices, as do some other specifications. Hence, using such a model one can synchronize several output modalities, such as picture and sound.
State of the art interaction requires compound hybrid devices. Each subdevice in the compound device must be capable of controlling a portion of the I/O streams. Moreover it must be able to accept externally provided data which may influence the way it behaves. For instance, a speech recognition system which uses a dictionary to enhance reliability must be able to use locally provided dictionaries. A real-time 3D scene renderer must accept view directions from an input device controlling the viewpoint. The movements of a vehicle in a 3D simulation might be controlled by spoken commands. The vehicle to be controlled in this manner might be the one that was selected by focussing the view on it. If I do not have a voice system available, it should be replaced by another option (e.g. dials) without having to change the underlying program. Perhaps in this case the speed of the vehicle would automatically adapt to a less efficient control technique.
In order to enable user friendly and efficient interaction, the optimal context must be generated. The mechanism for context realization is also based on controlling the interaction device behavior through impulses from the context. Whether this is done implicitly, e.g. by caching techniques, or explicitly, by making the context a visible part of the presentation, should be the decision of the interaction designer. If they want to reuse an interaction device in a different context, they should be able to do so without reengineering. A special case of this occurs when higher-level interaction devices determine the context of the lower level subdevices. Similarly, the other way around it should be possible to adapt the context to the interaction to short or long term history. Thus context impulses are bidirectional.
It is possible to extend current reference models such as PREMO in this direction. One goal of further work should be to clarify to what extent this is worth doing in a standardized fashion. It should be noted that this way of looking at interaction devices would widen the horizon for new devices as well as new combinations, e.g. graphics and image processing.
New Application Classes
A simplified characterization of the external behavior of a classical interactive application is to present the results of an ongoing calculation and to accept user input to influence further calculations. If the reaction of the system is real time, we call this computational steering. Steering is the current state of the art in I/O intensive interactions. It has revealed the power and potential of giving the user the sensation of being immersed in the application environment. Current applications are mostly concentrated at the high end of computing, such as large-scale engineering applications incorporating simulations. The resulting immersion provides a more efficient transfer of information between the user and the system. This is the reason why simpler forms of immersion should be brought to the desktop and home PC.
Some examples of application classes where interactivity can be improved are:
- Navigation through and updating of large information bases, where part of the interaction resources are used to maintain the user’s filter of interest. Such information systems require built-in links to presentation and navigation services that can accommodate a wide variety of platforms. A specific example is a Web browser that can connect with any local interaction device(s).
- Person to person communications, where all participants dwell in the same information domain.
- Distributed interactive information systems, where the resulting interaction is composed from the interaction facilities of each subsystem and is transparent to the user.
Many more examples can be given, but the above serve as good indicators of the massive scale on which improved interactivity can be applied.
New Features
Although some of the new features anticipated have been suggested earlier, here is a longer (yet certainly non-exhaustive) list of those deserving further study:
- Dynamic context: Intensive interaction can be streamlined with appropriate contextual help. However the stronger the context, the better the explanatory features in the context must be.
- Behavior rather than predefined dialogue: Dialogue scenarios become dynamic and non-deterministic because the interpretative rules are distributed over independent agents each of whom may trigger a reaction independently.
- High degree multimodality: Combining media both on the input and output sides, and merging them into hybrid devices.
- Recognition process as a part of the input mechanism: This is possibly one of the most significant technological improvements that must be enabled and become economically viable. The raw technology is available, but current software and hardware platforms are not yet ready for it.
- Automatic tracking and recording: Intensive interaction sessions create many reasons for tracking and recording. These include reuse, taking a second look at some significant phase, and remembering points of interest. Implementation implies that daemon processes accompany interactive sessions, making themselves visible on demand.
- On the fly assistance: This can be thought of as an on-line spelling checker for complex interaction entities.
All of these features can be realized through forms of external and internal input/output relations.
Why a Standard?
There are several reasons why a standard is desirable:
- Users will benefit from a competitive market where they can choose their platform from the offerings of competing suppliers and still be able to run highly interactive applications without excessive investments in custom interaction devices or in learning new techniques.
- A user should be able to connect the interaction device repertoire of a given platform to the “interactors” of plug-in applications. Moreover, several such interactors should be able to be combined to create the interaction setting for the platform.
- Efficiency requires that interaction combine seamlessly with compression techniques. This may put another burden on the shoulders of compression experts.
- A user should be able to personalize his or her interaction options and thus create optimal personal ways of working.
- For our discipline to progress, we need to consistently move from the level of abstract composite values to the level of processes that stand for dynamic values, i.e. to a new level of abstractions.
Dr. Paul ten Hagen
CWI
Amsterdam, The Netherlands
Difficulties to be Expected
The basis for this future standards work must be an agreement on a taxonomy of interaction, leading to a common class of basic abstract objects and methods. Currently, all individual application examples are ad hoc implementations. After we have such a taxonomy, the next step is for wide acceptance to be obtained. We should propose the reference model first, and then proceed with the next steps, such as a RFT or NP.
Some of the difficulties that must be overcome are:
- Unfortunately, the output first approach that has resulted from recent advances in presentation technology has created many entangled complex output processes that are insufficiently prepared for interaction. For instance some of these processes cannot easily produce the intermediate results required for relation to an input entity.
- Operating system and windowing support must become more open to real time and resource requirements. For instance repaint policies must become more consistent across systems.
- We need to find generic interfaces to semantics or contents that can filter out relevant data, e.g. a search engine embedded within an interaction handler.
- Strong, intensive interaction tends to break up the application in more fragments. A way must be found to add structure and consistency to such fragments.
- The role of cognition in interaction must be better understood so that it can be stated in terms of design principles.
- The process of developing an interaction standard along these lines must be made attractive enough to industry, such that a sufficient number of companies will see enough profit potential to justify participation in the effort.
- Tightly coupled I/O as foreseen here is even on the conceptual level understood only by a small number of experts.
Conclusion
This discussion has raised more questions than it has answered. My last question: Will we start working on interaction now or shall we leave it to another century?
Synthetic Environments
George S. Carson
GSC Associates
Scope
A synthetic environment is a representation of a natural or artificial environment, whose components are located in reference to a well-defined spatial origin and coordinate system. For representation of the natural environment, this origin is generally either the sun or earth, however any celestial body (or even imaginary body) can be used. An important special case is that of geo-spatial locations referenced to the earth (including its surface, oceans, atmosphere and near space.) A synthetic environment includes terrain and terrain features (both natural and man-made); models of objects such as avatars and certain localized features of the environment with complexity in structure or behavior (e.g., vehicles, buildings, smoke plumes and tornadoes); the ocean (both on and below the surface); the ocean bottom including features (both natural and man-made) on the ocean floor; the atmosphere including environmental phenomena; and near space. In addition, a synthetic environment includes the attributes of the objects in the environment, as well as the relationships among the types of objects. The representation includes constraints necessary to ensure correct automatically generated behaviors. These include topological and rule-based constraints. The purpose of the Synthetic Environments Working Group of the IGBT is to address all aspects of support for the creation, authoring and interchange of synthetic environments. The scope of the working group includes, but is not necessarily limited to, support for the creation of applications in these areas:- Simulation and modeling, including simulation based design
- Entertainment, including single user and multi-user games
- Instruction and training
- Media (versus, for example, “graphical environments”)
- Technology (for example, “world” is a VRML specific term for a similar concept)
- Representational polymorphism
- Conceptual models of the represented objects
- Data interchange formats
- APIs for reading, writing and/or accessing synthetic environments
- Real-time interchange of environmental information
George S.Carson
GSC Associates, Inc.
5272 Redman Road
Las Cruces, NM 88011
Tel: +1-505-521-7399
The copyright of articles and images printed remains with the author unless otherwise indicated.
Goals and Deliverables
- Determine if it is feasible to adopt international standards in the area of synthetic environments. This includes determining:
- The maturity of technical approaches to synthetic environments
- The degree of international consensus on the best technical approach(es)
- The benefits (or detriments) of standardization in this area
- As with all IGBT work, the plan is wherever possible to:
- Identify and harvest work already proven in commercial practice outside ISO and IEC
- Transpose that work into international standards making only necessary and mutually agreed changes
- The working group will determine the feasibility of using the proven work on the SEDRIS (Synthetic Environment Data Representation & Interchange Specification) project as a basis for one class of standards (non-real-time interchange among authoring systems) to be proposed by this working group. For more information about SEDRIS, you may visit the SEDRIS web site.
- Identify all other feasible base documents and approaches (if any exist).
- Write one or more New Work Item Proposals (NPs) for standards development to support synthetic environments.
The work of the Synthetic Environments Working Group will be completed in 1998. If possible, a New Work Item Proposal will be forwarded to JTC 1 for ballot. The leader of the Synthetic Environments Working Group is Karen Williams of the U.S. Government’s National Imagery and Mapping Agency. You can contact her for more information: kwilliam@MSIS.dmso.mil, or visit the working group’s www site.