ITEC 2000 Abstracts

Advances in modern technology have heralded a plethora of elegant solutions to both existing and emergent engineering problems. Increased automation in systems control and condition monitoring has been accompanied by a growth in the market for computer-based simulation and training. Such training can normally be shown to be highly effective, often supplying improvements in operator efficiency and allowing significant reductions in the overall time required to learn a particular skill. Arguing the case for investment in sophisticated training packages can routinely be supported by hard statistics; the benefits of investment in education, however, are less tangible. In the nuclear power industry for example, it has long been accepted that education acts as an insurance policy against undocumented occurrences; but without clear evidence of its worth and in the face of increasing budgetary pressures, it remains vulnerable. In this paper, the authors summarize their continuing research into the use of the concept of annualized probabilistic cost as a means to quantify the benefits of education, with particular reference to industries where the probability of an accident may be low, but the consequences may be devastating.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

QUANTITATIVE TRAINING SYSTEM ASSESSMENTS USING GENERAL SYSTEM PERFORMANCE THEORY”

Prof. George V. Kondraske,

Prof. Louis J. Everett, Assoc. Dean,

American University of Sharjah, School of Engineering

Mr. Robert Houser, Microsoft, Inc.

Ms. Sujatha Kashyap, Prof. Richard A. Volz,

Dept. of Computer Science, Texas A&M University

Ms. Mo Zhou,

To date, assessment of training systems is performed using complex, empirical, ad-hoc methods. This makes it difficult to compare one training system against another or to predict the training value added of different approaches or subsystem components (e.g., use of 3-D graphics instead of 2-D graphics in a training simulation). This paper discusses how we have approached the assessment process using a new framework called General System Performance Theory (GSPT). GSPT uses the notion of performance capacity envelopes defined by different dimensions of performance (e.g., speed and accuracy) as a basis for modeling system performance. In addition, resource economic principles are used to explain system-task interfaces. Thus, a system is viewed to possess "performance resources" (i.e., the envelope) and a task is viewed to make demands on these performance resources (i.e., as points in the space that define the envelope). We have used GSPT to develop performance models for the overall training system as well as for selected subsystems. GSPT was motivated by and is often applied to human performance modeling and measurement. As such, the trainee can be modeled using the same constructs used to model the training system. A particular example is its application to evaluation in the context of human rehabilitation. GSPT provides a quantitative basis for including even abstract concepts such as "user friendliness" into assessments.

The goal of our work is to use GSPT to develop the processes for quantitatively determining the value added by different training subsystem components without requiring the actual fabrication and testing of the training system. This will enable training system builders to ascertain whether or not it is worth the cost of incorporating different subsystem modalities or different levels of subsystem quality into their training systems. Ultimately, we foresee computer-aided training system design tool. To demonstrate how the process can be applied to training systems and to experimentally validate concepts in a training context, we are building a virtual reality/simulation-based testbed. In this testbed, we have chosen to focus initially on tasks that stress a particular human capability called Situational Awareness. Situational Awareness is defined in terms of the capacity to be cognizant of and properly prioritize important information, usually while executing several simultaneous tasks. Tasks of this type are executed by pilots and air traffic controllers, for example. In our experimental set-up, we can vary parameters (e.g., display resolution) that determine performance capacities of selected subsystems that comprise our training system and also measure selected lower level performance capacities of trainees. GSPT is used to develop causal models that attempt to link subsystem performance to overall training system performance, as well as to define the approach to measurement of trainee performance in training tasks and in the "real world" task for which they are being trained. The training system and experimental "real world" system are implemented using VIVIDS and a link we have built to WorldUp and WorldToolKit to permit 3-D simulation with modeling tools which can be run on a wide variety of platforms.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

INDIVIDUAL AND TEAM PERFORMANCE ASSESSMENT IN NETWORKED SIMULATION TRAINING

Jos van der Arend and Roger Jansen

TNO Physics and Electronics Laboratory

Within the Netherlands, a number of companies and research institutes involved in simulator technologies has been co-operating in a project called SIMULTAAN. The SIMULTAAN project partners have combined their knowledge and capabilities to design and build a distributed training simulation system. The result of the SIMULTAAN project is a generic HLA-based simulator architecture, which will be the baseline for the partners' future networked training simulators.

In the SIMULTAAN architecture, the SIMULTAAN Scenario Manager tool controls and monitors the federates (simulators and tools). An essential part of the SIMULTAAN architecture is the SIMULTAAN Performance Assessment SubSystem (PASS), which advises the Scenario Manager in choosing the best scenario that leads to achieving the training objectives. PASS optimises the learning curve of the individual trainee and the team.

The SIMULTAAN PASS component takes care of automatic analysis and assessment of trainee and team performance. During and after a scenario execution the trainee and team performance are analysed and judged; based on this automatic judgement the progress is determined. Options for this progress are: (1) restart the same scenario or (2) start a new scenario with simpler, more complex or alternative settings or (3) finish the training program. The instructor can always override the automatic judgement results or scores to influence the individual and team assessment and progress.

The automatic performance assessment in PASS is based on a generic framework with scenario-specific actions and action-related judgement rules. For each trainee or team of trainees a list of expected actions is prepared. These actions are automatically checked off during the scenario execution when the action-related judgement rules are found true or correct. The action-specific score weight assembles the final scenario score. The instructor or training developer prepares the expected action lists, the judgement rules and score weights for each scenario.

The PASS concept has been implemented and evaluated for the alarm rescue team performance in the SIMULTAAN project demonstration. In this demonstration the team consists of three fire brigade vehicles and a co-ordinating control centre. PASS proved to be a valuable and flexible generic concept for automatic performance assessment in networked simulation training.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

USING PERFORMANCE MEASURE TO REFINE THE TRAINING PROCESS

MODELING, MEASURING AND EVALUATING TEAM TRAINING

THE BENEFITS OF REAL-TIME TRAINING FEEDBACK: DEVELOPING PERFORMANCE MEASURES

Ms. Alison L. Young

DERA, UK

This paper describes a current experimental programme investigating delivery of training feedback options. The research explores the benefits of real-time feedback, and the medium of delivery. The work is being carried out to investigate the suitability of Augmented Reality (AR) technology for an embedded training system, which places the trainee in their actual working environment. AR has potential for such a system since it enhances the real environment with additional information (Young, Stedmon & Cook, 1999).

In particular, the paper describes the performance measurement issues that arise when dealing with a complex decision-making task. Significant emphasis has been placed on the process measures of performance since it is important to know whether the trainee is achieving the correct outcome in the desired way. Measuring human performance in this way is to be of increasing importance in the future, given that advanced technologies provide the capability to monitor detailed processes, thus allowing the human trainer to focus on team aspects of training.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

THE INTEGRATION OF TECHNOLOGY BASED TRAINING WITHIN EXISITING TRAINING ENVIRONMENTS

Lieutenant Commander M N Bowden Royal Navy

Once the time consuming and expensive process of specifying, designing and procuring a Technology Based Training (TBT) solution has been completed, the package has to be integrated into an existing, or possibly a new, training environment.

TBT packages range from multi-million pound high fidelity simulators to simple Computer Aided Instruction (CAI) tools written in-house by instructors. In order to function efficiently and effectively, however, they all need not only to fulfil the training objectives (and to prove quantifiably that they do so), but also to fulfil the training design and quality control requirements extant within any verified and validated training environment.

In order to ensure that this integration and evaluation process takes place smoothly, it has to be taken into account early on in the project, and planned for carefully. Instructors, training designers and internal quality controllers all need to be involved in the process.

The main role of the Royal Naval School of Educational and Training Technology (RNSETT) is to identify, disseminate and assure educational and training technology best practice throughout the Naval Service. On behalf of Flag Officer Training and Recruiting it undertakes regular quality audits of all Royal Naval training establishments, and has become increasingly involved with the introduction and subsequent evaluation of a varied range of TBT into the Royal Naval Training System. This paper will outline the key considerations that need to be made before any TBT product is delivered, and will propose a 'best practice' approach to integrating it into a training environment and subsequently evaluating it.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

COOPERATIVE COMPUTER BASED TRAINING- A NETWORK BASED TRAINING SYSTEM USING VR COMPONENTS

Horst Kramer

AIM Gmbh

The Co-operative Computer Based Training System (CoCBT) TeamTraining Amphibious vehicle M3" is probably worldwide the first training system that combines a tutorial concept with virtual reality (VR) and simulation components, to train teams in working on complex processes.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

UNIT BASED TRAINING (UBT) CONCEPT FOR DELIVERING THE UNIT LEVEL TRAINING FOR THE 21ST CENTURY UK ARMY

By M Kelly Sc(DBE).

The author wishes to acknowledge the help provided by Mr M Niven DERA (UK), Lt Col M Milligan SO1(Dev) DGD&D (UK) and Lt Col J N Morris DBE(Sim) (UK).

The delivery of Collective training in the UK Army has been well served recently but Unit training is still in the sand table and black board era. A need has been identified to deliver Unit Based Training (UBT) in a coherent modern and cost effective form. To support this the UK MOD Applied Research programme is constructing a Technical Demonstrator Programme in partnership with industry to validate and define the concept for the Unit Based Trainer. This will capitalise on existing infrastructure and expand and develop the Army training management strategy. It will benefit from the commercial trends in training delivery and the new power of distributed PCs, the WWW, CBT and PC games.

The programme will examine the architecture, software standards, training management, course authoring, training delivery, performance assessment and record keeping of a distributed UBT system. The development of structured training schemes, assessment systems and interfaces to Synthetic Environment based training will be researched.

The benefits to military training will be to increase match to and efficiency of Collective Training, to train Collective Performance levels 1-3, to compensate for identified skill fade in digitized skills and to optimise Territorial Army training. UBT will provide training in theatre, on route (ship), at home, by asynchronous, self paced, assessed, personalised training; indeed training 'anywhere anytime'. Team, international, and combined training will be addresses at the Unit level. UBT will provide an audit trail for training to defend against potential litigation

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

PLANNING AND ANALYSIS FOR WEB BASED TRAINING IMPLEMENTATION

Deborah A. Williams, TRW

Distance Learning Group

While the benefits of adapting a Web Based Training (WBT) approach are becoming widely understood across private industry and government arenas, too often the planning and analysis which will facilitate an effective WBT experience is overlooked. This paper will focus on the necessary planning and analysis that should go into the design of WBT products as well as the critical administrative and management infrastructure which supports the WBT project. The paper will provide information relative to:

* Organizations seriously considering a move to WBT will gain an understanding of the critical planning and analysis they must engage in to ensure the success of a WBT initiative.

* Steps inherent in designing a WBT system, which achieves long term objectives for the organization, will be explored. Special emphasis will be placed on anticipating long-term needs and incorporating this into the WBT infrastructure and design.

Internal and external factors influencing WBT design and implementation decisions are discussed from a "lessons learned" perspective which has been gained from five years of experience in the design, implementation, delivery and maintenance of enterprise level WBT systems for large government, military and private organizations.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

INVENTIVE THINKING (IT) - A GIFT OR A SKILL ?

Moshe Ben-Porath

MHT General Manager - Israel Aircraft Industries

The paper claims that inventive thinking can be developed using specific methods and tools thus introducing in the organization a culture of inventive solutions to problems.

The hierarchy of knowledge: Data, Information, Knowledge, Wisdom and Ingenuity, was previously defined in a paper given by the author at ITEC 99. To implement IT as a routine, one should first reach the level of Wisdom on the Intellectual Ladder and climb to the level of Ingenuity. The highest level of Ingenuity is IT. The preliminary need before concentrating on the methods for IT development is the enhancement of generic thinking skills, i.e. development of cognitive parameters. This is done at MHT, the training division of Israel Aircraft Industries (IAI), by integrating the Knowledge Acquisition methodologies developed by MHT, with Instrumental Enrichment tools developed by Prof. Feuerstein. This integration supplies the trainee with the ladder to climb from Knowledge to Wisdom and from Wisdom to Ingenuity. The paper presents the principles of IT and the methodology of TRIZ, developed by Genrich Altshuler, which supplies the trainee with a technique that enables the implementation of IT to solve problems. This paper presents a solution implemented in IAI by a team that was exposed to the IT methodology. The TRIZ theory defines five levels of solutions: 1.standard 2.improvement 3.invention inside a paradigm 4.invention outside a paradigm 5.discovery. Whenever a person faces a technical contradiction, the common engineering approach is to design a compromise. The TRIZ IT does not accept this approach and instead leads to an invention which surmounts the contradiction. The methodology is leading the person to convert the technical contradiction into a physical contradiction where the contradiction lies within one physical parameter that should change in opposite directions. Then the person solves the contradiction by climbing on a ladder of abstraction.

The IT method, together with software tools, integrated with IE exercises, enable a safe climb on the intellectual ladder up to IT, thus enhances the Intellectual Capital of the organization.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

WEB BASED TRAINING AND SUPPORT: BEAMING UP SCOTTY

Ms Monique M. Volant

SOGITEC Industries

Mr Filippo de Stefani

Alenia Marconi Systems Radar Division

Mr Gregory O'Hare

Department of Computer Science, University College Dublin (UCD), Belfield, Dublin,

Mr Stefano Susini

University of Siena - Telecomunication and Telematic Laboratory

ECHOES, Educational Hypermedia Online System (a EU Educational Multimedia programme) is a distributed simulated environment designed to provide training and support to field engineers. Using Computer (Web) Based Training, Virtual Reality and Intelligent Agents techniques, ECHOES facilitates courseware delivery and computer supported co-operative work.

Engineers connecting to the ECHOES system via their computer, on Internet or an Intra/Extranet, encounter a user interface based on the visit metaphor and are presented a virtual environment which they are free to explore. Within this environment, different services are available: training, chat, library, together with troubleshooting support so that the user can take advantage from the experience of other technicians and experts.

Benefits expected are twofold: the time required for classroom training is reduced and engineers can get a lot of information from the system without necessitating human support.

After a brief introduction to the ECHOES Consortium, the paper presents how the system has been designed and how it can be used to provide periodic training refresh and support to field engineers. Preliminary results of user trials are described and commented upon.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

IMPLEMENTING VR TECHNOLOGY TO MARITIME AND OFFSHORE TRAINING

Pertti Broas

VTT Manufacturing Technology

Virtual prototypes are used mainly in the design process, where systems can be tested in virtual environment before they are built. This helps the designer to optimise the product better than earlier especially what comes to human interfaces. For example ergonomics and workplace layouts can be tested for different sized users.

The same virtual models can be used in training when the trainee can familiarise himself and practise with a computer using a virtual product in virtual environments. This kind of training helps both the designer and the trainee.

This paper discusses the use of virtual technology applications in maritime and offshore areas which are under development in a ESPRIT research program DISCOVER. The prototype scenarios already demonstrated include incident management training in the case of fire in the engine room onboard a cruise-liner or in accommondation spaces on an offshore-rig. Virtual modelling of incident, functionality of the simulations and communications between participants are handled using a multi-user simulation in a computer network.

This kind of training is based on the growing needs of end-users - for example companies operating in production, integrated transport (port operators, shipping agents, road haulage and shipping companies) and maritime sectors and the training providers serving them. Training with virtual prototypes will ensure that their working procedures are efficient and safe and that their employees consistently perform competently.

Increasingly complex working environments (e.g. ships' bridges and engine rooms, offshore and onshore control rooms) coupled with reductions in manning levels, new organisational structures and increased use of technology are placing significant demands upon individuals and teams to learn new tasks and procedures whilst becoming even more reliant upon their working colleagues to perform competently.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

EMBEDDED STIMULATORS FOR NAVAL ON-BOARD TRAINING

Mr. James R. Cooley, Principal Engineer

AAI Corporation

In many navies, readiness training is moving out of the classroom and onto combatant vessels. This evolution is being driven by two factors: declining defense budgets and the need to improve readiness of combat system team crewmembers. In the past, combat system team training has consisted of two major elements: classroom training and underway training exercises. Both of these training venues come with a host of disadvantages.

Classroom training requires the maintenance of an extensive shore-based training infrastructure; the actual training value is limited by the fidelity of the training equipment. Shore-based trainers typically include equipment and sensor models that only approximate what the combat system team will find on real ships in a real combat environment. This is due to the nature of the shore-based facility; shore facilities must serve as trainers for many ship classes, and shore-based trainer mockups tend to lag the state of the actual combat system equipment by several years.

Underway training in a canned warfare environment with real participants has the advantage of training occurring on real equipment in a real situation. The expense of these exercises, however, severely limits their utility and the degree of realism that can be attained. Declining defense budgets have also severely limited the opportunity to train in this environment.

The alternative training strategy is a blend of classroom trainers and shipboard exercises. An on-board training stimulation system is provided that presents a synthetic training environment to the combat system team on real equipment. This approach combines the best of both worlds. A stimulation system can synthesize complex combat scenarios much like those possible at a shore site and provide training on actual equipment just as sea training exercises do. Furthermore, an on-board stimulation system is available at all times to the crew, even at sea; therefore, training opportunities abound.

The design of an on-board training system is built around a number of building blocks that are dependent on the type of equipment included in the host combat system. Stimulators and simulators are provided for each element in the combat system and are networked together and controlled from a simulation computer. The design of the network that connects the boxes together is such that any complement of boxes can be included without altering the software that executes in the simulation computer. This design is achieved by using a DIS/HLA protocol for the network architecture and providing a DIS/HLA front end to each of the stimulation or simulation boxes. The on-board training system is configurable to any combat system by selecting which stimulators and simulators are needed in a particular situation.

This paper describes the design of a representative on-board training system with emphasis on the radar stimulation building blocks.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

THE INFLUENCE OF STCW 95 AND THE ISM CODE ON THE EFFECTIVE USE OF SIMULATION

EMBEDDED TRAINING: POTENTIALS AND CHALLENGES

Hilbert Kuiper, Anja van der Hulst, Geert Slegtenhorst

For years Embedded Training was not a fully-fledged alternative amongst other training media such as distributed simulation, simulators, electronic classroom, and part-task trainers. However, technology is rapidly evolving, which manifests itself for instance in far reaching capabilities in the field of telematics. Parallel we see a larger mobility of people, and increasingly unpredictable scenarios. These developments urge for just in time learning, learning on demand, and integration of working and learning. As a result, there is nowadays much more focus on the possibilities of Embedded Training. Embedded Training coincides with the tendency for Electronic Performance Support Systems (EPSS).

Embedded training is defined as a capability built into or added onto operational (sub)systems to enhance and maintain skill proficiency. Literature distinguishes three types of Embedded Training: (1) appended, defined as an add-on to the operational system and with some built-in components in the operational system, (2) umbilical, the same as an appended system but with a physical connection to external systems, and (3) fully-embedded, defined as fully integrated with the operational system and having the possibility to connect to other training facilities at location or at a distance.

We will focus on the potentials and challenges of the fully-embedded type, the spectrum reaching from operator training upto and including integration with Command and Control.

The presentation will take as a red thread running through it depicting how a future (2010+) maritime training scenario looks like using the potentials of embedded training and other training facilities, i.e. towards a common synthetic training environment for joint/combined operations in which embedded training plays a significant role. In this scenario, training facilities on-board as well as ashore will be linked and use will be made of state-of-the art and emerging technologies (e.g augmented reality, speech recognition and synthesis, intelligent trainee-system dialogue, tele-coaching) for embedded training applications.

Using this scenario we will focus on the implications and challenges of embedded training as it can be seen at three different levels: (1) organisational level, (2) technical level, and (3) pedagogical level. Examples of questions that can be posed at the different levels are:

At the organisational level: how to organize new training programs, how to fit embedded training in the program to make a seamless link to other training means, what are the personnel implications, does it include a complete organisational change with respect to training and education?

At the technical level: what are the implications for the acquisition of an operational system, i.e. how to guarantee the specification of embedded training requirements from the very beginning; what are the potentials of new technologies, e.g. using augmented tactical displays for real-time performance feedback?

At the pedagogical level: what to train embedded and what not, what to training on-board and what ashore, what learning goals can be reached with embedded training, how to structure the training, how to test the readiness, how to obtain performance measurement data?

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

MULTI-SHIP TRAINING FOR THE TOW-OUT OF A LARGE SUBMARIINE

AIR FORCE MODELING, SIMULATION AND ANALYSIS EDUCATION AND TRAINING TIGER TEAM”

Dr. Mary C. Fischer, Executive Director

Air Force Agency for Modeling and Simulation (AFAMS)

Dr. Ronald W. Tarr, Program Manager

Performance Technology Group

The Air Force Modeling, Simulation and Analysis Education and Training Tiger Team (ETTT) is an effort to identify the core body of knowledge required for all Air Force personnel newly assigned to modeling and simulation positions. On 15 Oct 98, the Air Force Modeling, Simulation and Analysis Working Group (AFMSAWG) approved the establishment of the tiger team to perform an in-depth study to determine AF specific requirements and to identify AF solutions. The ETTT will identify educational opportunities and core M&S education requirements, leveraging existing efforts where possible, such as the Defense Modeling and Simulation Office (DMSO) M&S Education Program with the goal of establishing a state-of-the-art, cost-effective solution.

To facilitate this effort a joint team was established composed of AFAMS personnel augmented with researchers from University of Central Florida. The team's objectives are to examine performance requirements across the full spectrum of Air Force Modeling, Simulation and Analysis, develop core M&S competencies, identify existing education and training opportunities, and recommend an approach to meet recognized shortfalls. This paper will discuss background, structure, methodologies, products, and follow-on effort.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

MULTIDIMENSIONAL TASK ANALYSIS

William A. Platt, Stephen J. Guynn, Ronald R. Rising

Increasingly authors are noting both the importance and difficulties associated with task analysis performed as a component of training analysis leading to the design of instructional activities and materials. Difficulties can arise when a task analysis approach is used in an inappropriate situation. Different approaches yield different types of data. Traditional behavioral approaches tend to describe observable activity. Newer cognitive approaches focus on knowledge content, decision tasks and rules. To offset the difficulties inherent in each of the many separate methods and models of task analysis, the authors have combined a number of methods each designed to harvest data in a different dimension of the job-task setting.This paper reports on the use of multidimensional approach in the Veterans Benefits Administration to analyze the newly created job of the Veterans Service Representative (VSR). Traditional task analysis methods were used along with participant observation to establish the relative importance of salient features of the job task setting. The relevant dimensions are (1) Organizational design, (2) Work Breakdown Structure, (3) Knowledge Content, (4) Work Process Flow Logic, (5) Typical Events Time Line, (6) Communications Network, (7) Physical setting Traffic Flow and Perceptual Envelope. The analysis then continued using models appropriate to each dimension. The analysis models were obtained from the literature or created by the analysis team. The results yielded a more holistic view of the job task and enabled the team to make distinctions and conclusions that will be helpful in designing instructional sequences. The fundamental structure of the job defined by law and court decisions could now be contrasted with the subtleties of practice identified using the many sided view. This included a great deal of heuristic activity and variation from office to office. Insights relevant to training design were obtained by comparing each data set. A multiple approach allows findings to be compared by mapping results form one to another. This also provided a stimulus for the investigation of missing data. The authors believe that this approach is worth repeating and have made suggestions for additional dimensions to be examined.

Conclusions: The multidimensional approach seemed to fit the situation presented in the VBA engaged in business process re-engineering, training development, and down sizing. The mix of descriptive, behavioral and cognitive approaches was useful in such a dynamic and changing situation. The-cross checking made possible by using the various views, did reveal aspects of the situation that will be important to developing instruction. The concept of core tasks surrounded by a local office supporting shell of implementing tasks evolved only after the multiple views were clear. Thus it was recognized that training must focus on both aspects of the job. Use of a multidimensional approach is recommended for task environments where active change is the rule, and for use in professional settings.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

TASK ANALYSIS: FRIEND OR FOE?

STRUCTURING A SIMULATION EXERCISE DATABASE USING STREX

Michelle Joab, Odette Auzende, Michel Futtersack, University Paris

Patrice Le Leydour, THOMSON TRAINING & SIMULATION

When initiating a training course, an instructor is faced with the complex task of retrieving the appropriate exercises and building the relevant sequence (a teaching module) to help the trainees gradually acquire a given skill, either as individuals or as a team.

Retrieval of an exercise is a complex task because most of the useful selection criteria (e.g. the teaching goals) do not explicitly appear as part of the data set that makes up the exercise. Available exercise data mainly refer to parameter values that the simulation software will further manipulate in real-time. Furthermore, lack of information about the exercise complexity level makes it difficult to arrange exercises in a structured sequence.

The availability of a set of training aids that will enable a population of instructors to characterize existing exercises and further extract structured teaching modules therefore represents a significant step forward compared to today's instructional environments.

This paper briefly describes STREX (STRucturing EXercise database) as a generic approach to answer this kind of instructional need. Using examples from both TRUck Simulator for Training (TRUST) and Crew Training Simulator for the LECLERC battle tank, it describes how instructors are given the possibility to add new selection criteria to the existing set (possibly dependent on the previous ones), visualize the exercise database contents using a tree-like representation and generate teaching modules according to the trainees' needs.

STREX is a tool that is independent of any training domain and can be connected to any exercise creation tool; STREX software integration simply requires the importing of the exercise data into the STREX database. In order to generate a teaching module, the instructor has to define the number of exercises, set the desired selection criteria and their relative importance and finally characterize them by entering trigger values. STREX will then generate a balanced module meeting these constraints; the instructor will remain free to modify the balance of the module as well as its contents.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

A STUDY OF SUBJECT MATTER EXPERT REQUIREMENTS FOR SIMULATION INSTRUCTORS AND HUMAN FACTORS ENGINEERING OF SIMULATION OPERATOR STATIONS

ANALYZING TRAINING RESULTS USING DATA MINING TECHNIQUES

Dr. Amnon Gonen, Operations Research Director, GWG - Global War Games Ltd.

Inbal Ben-Israel, Senior Software Engineer, GWG - Global War Games Ltd.

Modern computer data mining systems self learn from the previous history of the investigated training system, formulating and testing hypotheses about the rules that this training system obeys. When concise and valuable knowledge about the training of interest had been discovered, it can and should be incorporated into some presentation and lesson learning system.

Training simulators usually poses an After Action review (AAR) module that assists in learning from the previous training sessions. The AAR does not handle the analysis of several training sessions, past training sessions, comparison among training sessions etc.

The data mining is widely used in civilian applications and here it is first introduced into a training simulation, looking for statistical connections among the results of training sessions.

The methodology used in this study is as follows:

Detailed description of the database to be analyzed. In this case, the database was "artificially" generated to reflect all kind of relations between the attributes.

Database preparations. This task is very fundamental and includes merging of tables, manipulating records from several training sessions, manipulating fields to define new attributes, visualization of the database and data checking to ensure the necessary quality of the database.

The next step is modeling and deriving results from the data mining software.

Examples of interesting results and limitations of the different algorithms are finally provided.

* The hit rate in engagements and the parameters influencing it.

* The connection between the impact location and other parameters.

The summary presents the main lessons learnt on the usage of data mining for debriefing training sessions.- -

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

DATA PARALLELISM VS. CONTROL PARALLELISM IN SYNTHETIC ENVIRONMENTS VISUALIZATION

Prof. Felicia Ionescu, Assistant Andrei Jalb_

Politehnica University of Bucharest,

The paper is dealing with the problem of synthetic environments visualization in shared memory multiprocessor systems. Multiple processes (or threads), executed on available processors of a multiprocessor system, partition the traversing operation of the scene graph representing the synthetic environment and communicate with each other through shared memory. As usually in parallel processing, two approaches can be applied to the partitioning problem of the scene graph traversing operation: control parallelism and data parallelism. In control parallelism approach, the scene graph traversing operation is divided into a sequence of stages, each of them executing a subset of the operation set needed in each node of the graph. Each stage of traversing operation is implemented in a process (thread) and processes are executed in a pipeline mode. This approach is very easy to implement, but has a number of limitations. The first problem with pipeline execution of traversing operation is the difficulty to balance the execution time of processes, and the worst case execution time of one process delays all other processes in the pipeline. Another problem is the limited number of processors that can be used, due to the small number of traversing stages.

Our work tries to compare pipeline execution of traversing operation with data parallelism approach. The scene graph of synthetic environment is an irregular structure, which cannot be statically balanced partitioned, and, much more, in different image generation frames, different sub-graphs are visible, depending on the observer position. Dynamically partitioning of the scene graph is a more suitable solution, which can exploit the full execution time of all processors. The state-space tree of graph traversing can be efficiently represented as a stack in which every unexplored path is stored at each step. Every entry in the state-stack stores the root of an unexplored sub-graph together with the complete traversing state (transformation matrices, materials, textures, light sources, etc), using adequate accumulation of states along the traversing path. The access of multiple processes to shared resources (state-stack and graphic interface) is synchronized using mutual exclusion mechanisms (such as mutexes, semaphores, events). Experimental results show the influence of scene graph layout on the parallel speedup. For the same number of polygons in the database, a significant speedup in data-parallel approach is obtained for a medium grain size of the scene graph and this speedup is greater then in pipeline execution.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

AUTOMATIC GENERATION OF 3D VISUAL DATABASES FOR SIMULATION AND TRAINING SYSTEMS

Izi Peled and Ran Yakir

BVR Systems (1998) Ltd.

A major cost-driver in modern day training and simulation systems is the visual image generation. Furthermore, the image generator (IG) is the principal output by which these systems are judged by users and potential customers. A smooth running high fidelity visual database will demonstrate simulation systems more effectively and enhance the overall training value as well as ensure system acceptance by end users.

Conventional methods of creating 3D visual databases depend on a large amount of manual labor in order to create a realistic and accurate training environment. This amount of manual modeling and production is prohibitive both in cost and in time-to-market. In addition to the requirement for realistic, high-quality visual databases in many applications, such as simulations used in training exercises, the real-time performance is critical. Consequently, a tradeoff is made and the database's visual quality is compromised in order to achieve the goal of high-level real-time performance. This paper presents an innovative approach developed by BVR Systems for automatic generation of high quality, real-time 3D visual databases. In addition, it presents real-time software solutions that were developed to further enhance the capabilities and fidelity of the visual databases.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

PC-BASED DYNAMIC TERRAIN SOLUTIONS

Mr. Graham Upton, Diamond Visionics

Over the last several years there has been a growing requirement for Ground-based simulation training systems. As part of this requirement there is also a need for added realism within the simulation to provide, in real-time, the manipulation of a simulated terrain database in a physically realistic manner during an interactive simulation. Dynamic Terrain is not new to the Ground-based simulation community. However, current technologies require high-end computational platforms, are not real-time (30Hz), and are often cartoonish in appearance.

This paper will examine techniques to provide real-time dynamic terrain in a commercial-off-the-shelf (COTS) PC with commercially available graphic accelerator cards. The task of developing Dynamic Tessellation is challenging, especially on a PC-based system. Dynamic Tessellation provides the ability to deform terrain anywhere in the database in real time without the need for predefined deformable areas. Both Pre-Tessellation and Instantaneous-Tessellation approaches will be reviewed as well as the effects of soil dynamics and dynamic texture.

Dynamic Terrain is a requirement for realism for the maneuver forces in the Synthetic Environment. Specifically the application of dynamic terrain encompasses mine breaching, bomb damage, building damage, soil plowing and snow plowing. Specific applications of dynamic terrain are for the Grizzly Trainer, the Armored Vehicle Launched Mine-Clearing Line Charge (MICLIC), Track Width Mine Rollers and Explosive Standoff Minefield Breacher (ESMB). Commercial applications include mining operations, air-traffic training for snow clearing and heavy equipment simulation.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

THE REQUIREMENTS FOR REAL-TIME IMAGING SENSOR SIMULATIONS USED IN MILITARY FLYING TRAINING

Mr Jon Platts, DERA Bedford,

Mr Frank Rutley, CSC Computer Sciences Ltd,

Mr David Green, CSC Computer Sciences Ltd,

The Defence Evaluation and Research Agency (DERA) is currently undertaking research sponsored by the UK Ministry of Defence into the Air Training Environment. The overall research package is focused on the technology and human factors issues which underpin all aspects of military training for airborne missions. Simulators will be used for a full range of training tasks from basic training involving the handling of a single air vehicle, up to mission rehearsal with many vehicles participating. Consequently the need for the correct and apposite simulation of appropriate sensor systems must be addressed to provide the complete environment in which to immerse trainees and exercise participants. To this end a joint study between the Flight Management and Control Department of DERA Bedford and Computer Sciences Corporation (CSC) has been completed that specifically examines the requirements of any imaging sensor simulation. This paper identifies the key features that need to be included in any imaging sensor simulation, giving the customer the ability to match precisely any future system requirements to a particular training need.

In order to provide the fullest answer to the question of how to relate sensor system simulation fidelity to training need, the team defined the training needs associated with sensor operation skills requirement. This was achieved by performing a skills and task analysis for representative aircraft and mission profiles and combining this with in-service training philosophies to produce an overall matrix defining minimum simulator fidelity requirement against the desired training outcome. Key practitioners from each major aircraft type were interviewed to discuss their experience of the actual method of use of the equipments. The paper is a synthesis of all the research undertaken at individual units. Common themes showing similar fidelity requirements for similar or disparate scenarios are drawn together and differences highlighted. Output is a pair of matrices showing mission type, training need and fidelity required for Imaging Infra Red (IIR) equipments and Night Vision Goggles (NVGs). The matrices provide information on the key features of NVG and IIR simulation for each of the operational roles, in priority order. The paper also discusses simulation to various levels of realism and provides a guide to the benefits of simulators on a sliding scale from full physics-based total system simulation, through partial realism, to low tech soft simulation of the sensor display.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

FACTORS TO CONSIDER WHEN BUILDING SYNTHETIC
ENVIRONMENTS WITH COORDINATED EW AND COMM THREATS

Edward V. Rivard, AAI Corporation

Over the years, many threat environments have been built for use in large, single purpose institutional trainers or as part of system evaluation exercises. Most of these threat environments have included traditional EW threats (Radar and IFF emanations), and some have included threat network Communications (both data and voice), but few environments have been built that combine the EW and Comm threats in a way that the two work together to produce an environment that recreates what is seen and heard.

Now that training and evaluation exercises are becoming larger and incorporating all kinds of disparate, distributed simulations, there is a need to create an integrated EW and Comm threat environment where coordinated activity can be presented for warfighter training and system evaluation.

This paper addresses some of the issues involved in making such a combined threat environment using reactive EW and Communications networks elements. It illustrates the importance of having both Comms and EW threats working realistically together in an exercise so the warfighter is presented with an environment where the contributions of each kind of jamming asset is accounted for fairly.

As an example of the features constructive EW and Communications threat models should have in this kind of cooperative simulation environment, a simple GCI (Ground Control Interceptor ) situation will be discussed. This case will be used to illustrate how disrupting the traditional EW (IFF and Radar emanations) and the traditional Comms Voice and Data links can have similar effects and how combining them provides models with both greater realism and wider applicability.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

CEPA 11 AND ITS STRATEGY

INTEROPERABILITY: THE EMPEROR'S OLD CLOTHES

Paul W. Sutton, Interoperability Manager

Advanced Technology and Prototype Systems (PD 13)

Space and Naval Warfare Systems Command (SPAWAR)

Differing definitions and views of interoperability have resulted in standards and interfaces that are not always compatible or interoperable with one another. Interoperability standards and interfaces do not necessarily provide the conditions necessary for real, end-to-end interoperability. Existing models of interoperability do not provide consistent, objective measures of interoperability performance. They do not generally describe interoperability in sufficient detail to understand how it really works or explain what causes interoperability performance to vary. This paper provides a conceptual framework of interoperability that addresses these deficiencies. It postulates a single, consistent measure of interoperability performance, such as the probability of successful interoperation, to quantify and understand interoperability. It also proposes a causal model of interoperability that is needed to determine which factors contribute materially to interoperability performance in order to make better use of increasingly scarce resources. Similarly, it presents a descriptive model to explain how interoperability works.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

SIMULATED GPS NAVIGATION IN HLA FEDERATIONS

Dr. P.C.A. van Gool

Illgen Simulation Technologies, Inc.

All modern flight simulators provide either specific or generic simulation of aircraft subsystems including models for engines, navigation and autopilot. These systems provide pilots and trainees with a computer-generated representation of the environment and accompanying dynamics. When it comes to GPS navigation however, most simulators use some form of pre-recorded data to simulate the operation of GPS receivers. There is a definite need for a more precise simulation of GPS navigation.

A technical framework has been developed by the US DoD which is to be applied to the whole range of potential modelling and simulation applications in the DoD. The framework has also raised much interest in the civil market. The architecture that is part of the framework, High Level Architecture (HLA), will be the common framework used by those simulations.

The paper will describe the efforts made by ISTI to provide simulated GPS navigation in HLA federations. The Global Positioning System (GPS) is an all-weather space-based navigation system to satisfy the requirements to determine accurate position, velocity, and time in a common reference frame, anywhere on or near the Earth on a continuous basis. The system consists of three segments. After a discussion of each of these segments, the paper will proceed with a discussion of the HLA and its potential benefits for re-use of simulation components. The paper will continue with a discussion of the global design of the simulated system. The paper will end with a discussion of the benefits of the current system and recommendations for future work on expanding the system's functionality and applicability.

The developed software subsystem improves the simulation, and therefore the training experience, in various ways. Introduction of errors into the GPS satellite constellation or introduction of various atmospheric effects are experienced by all receivers in a consistent way. This enhances the degree of reality and fidelity of each individual simulation and or the distributed simulation as a whole.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

DIS NETWORK TRAFFIC PREDICTIONS FROM EMPIRICAL DATA

Dr. Peter Ryan and Dr. Lucien Zalcman,

Aeronautical & Maritime Research Laboratory,

Defence Science & Technology Organisation (DSTO),

Distributed Interactive Simulation (DIS) defines an infrastructure for creating a synthetic environment by linking simulations of various types at distributed locations. DIS operates through broadcasting information about the simulated entities via Protocol Data Units (PDUs) across a network. For exercises with many simulated entities emitting PDUs frequently, this broadcast paradigm can result in considerable network traffic. Standard models predict that DIS network traffic will increase linearly as the number of entities in the exercise increases. However this assumes that each entity will participate at the same level of activity and ignores the complexity of warfare interactions. In this work, a series of experiments was carried out to measure the DIS traffic issued from standard platforms using the synthetic environment system ModSAF and its derivative NavySAF. Data were measured for scenarios with computer generated tanks, ships, helicopters, and fast jets and also for manned asset simulators which typically generate higher traffic rates. The results showed that the assumption of linear scaling for network traffic holds for air and maritime entities engaging in combat beyond visual range. However, for land entities such as tanks, the traffic scales non-linearly except during quiescent periods of activity such as transiting to a waypoint. During periods of intense engagement network traffic for land warfare increases in an unpredictable manner with consequences for simulation network development. For aircraft and ships, weaponry, such as missiles and close-in-weapon-systems, can generate extremely high traffic rates for short periods of time. These high traffic rates must be taken into consideration when designing experiments and simulation networks.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

NEXT WAVE SOFTWARE TECHNOLOGIES AND THEIR IMPACTS ON MODELING AND SIMULATION

Tom Strelich, tstrelich@illgen.com

Illgen Simulation Technologies, Inc.

The recent introductions of Sun's Jini1 (pronounced "Genie") technology and the World Wide Web Consortium's (W3C) eXtensible Markup Language (XML) have the potential to fundamentally change the way software is developed, deployed, and utilized since they represent the next logical step in the progression and fusion of distributed component architectures with network technology. The paper describes the problem domain and new software technologies, assesses their impact on simulation integration and interoperation, and provides a representative scenario illustrating the application of the technologies in a prototype simulation integration architecture.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

A COMMON ARCHITECTURE FOR CREW SKILLS & TACTICAL TRAINING SYSTEMS

Eytan Pollak, Ph.D., Lorie Ingraham, Senior Software Engineer, Mark Falash, Software Engineer

Lockheed Martin Information Systems,

Simulators for military aircraft and ground systems have been in use for almost two decades. During that time, simulation has proven to be successful in training individual, crew, and collective tasks needed to function in real aircraft or ground vehicles. Traditionally, individual crew skills trainers (such as gunnery) and tactical / battlefield trainers have been designed and implemented independently. Skill-based crew training systems have been developed as stand-alone systems or as a small group of tightly coupled systems with pilot, gunner, driver, or other highly specialized environments that have required highly interactive instructor/operator stations (IOSs). Tactical training systems have evolved somewhat differently and are designed to communicate entity state, tactical environment, and other data over a distributed network that is monitored to collect, assemble, and present training After Action Review debriefings for those being trained. Greatly improved training benefits and life-cycle cost benefits could be realized by developing a common architecture that allows both specific crew skills and collective tactical tasks to be trained using a single simulation system. The need for separate simulation systems was generated due to the different fidelity requirements of the hardware and software components of these two types of systems. Tactical simulators have generally sacrificed fidelity to gain scalability and to keep costs within affordable ranges. Skill-based crew simulators have required faster update rates, high fidelity, and high resolution. Also, tactical simulators have generally used free-range stochastic scenarios during training, while skill-based crew trainers have generally used well-defined and repeatable scenarios developed to resemble live-fire ranges and driving courses. For many years, these fundamental differences have driven the different architecture approaches used for these two types of training simulation systems.

In this paper we will present an approach to cost-effectively combine individual skills and collective skills training into one simulator set, analyzing and discussing the system architecture issues. We will present an infrastructure that can support both types of training by maximizing the use of common trainer services and components. A common architecture that allows the combination of these two systems into a single simulation system will be able to support both skill-based crew training and unit-level tactical training. This combined type of simulation system has been needed since the beginning of simulator-based training. Due to recent advances in architecture development techniques, computer hardware, and tools, it is now possible.

This paper is available on the 2000 ITEC CD-ROM. Order it from ITEC, Ltd. - Papers

AN HLA-BASED DISTRIBUTED SYSTEM INFRASTRUCTURE

Jean-Marc Naud, Makan Pourzandi, Ph.D., Yvan Lagacé

Virtual Prototypes Inc.

This paper describes a software product for simulation development and integration that provides a simple path to migrate existing legacy simulations to the HLA Run Time Infrastructure (RTI) and to build distributed simulation systems. It is aimed at existing owners of simulation applications interested in easily deploying their simulation in a distributed, HLA-c