(Eikenberry, 1999) An easy way to create metaphors to ease learning is to follow the 3C model.  Create – this is always a tough step in an endeavor, but you can help your creative ability by determining the part of a course you want to focus on.  Once you’ve determined what part you are going to work on, write down the general parts of the topic you wish to cover.  If you have small children or grandchildren, the next step may be easier.  Ask yourself how you would teach the topic to a 8 or 10 year old?  Write down all the ideas that come into your head, regardless of how silly they may seem. Go through the list and make associations with the list Connect – this is a quick step.  Compare the metaphor ideas with the content of the topic.  Compare a potential metaphor with the content.  Ask yourself are the elements connected? How are they alike? How are they different? Combine – determine how to combine the metaphor into your instructional flow.  Can it become a theme for the class?  Think about how to introduce it.  Can you use more than one of the senses to reinforce it?  Good Luck !

According to Coughlan (1995) a special purpose simulator is distinguished from a simulation language or system by the three properties above.  Most simulation systems are only used by trained experts.  The special purpose simulator is developed to insulate a user from the complexities of the underlying system. The main part of a special purpose simulator is the parameterized model of the system being studied.  The front end interface guides the user in the set-up of the parameters in a similar manner to many familiar software wizards.  These models have been created for many industries and for many purposes.  They are targeted at users with almost no simulation experience.  These include manufacturing engineers and managers.  One MS Windows based simulation package is SIMAN, a commercial spreadsheet, which uses BASIC as the underlying programming language.

Microsaint was developed in 1985 to fill in the gaps in simulations for human based tasks.  According to its developers, Microsaint fills in the gap in commercial simulators by targeting industries that have underutilized simulation.  These industries include manufacturing, healthcare, process reengineering and human factors  (Drury, 1996). Microsaint uses a graphical development system to ease development by users who are not computer savvy.  Its development methodology is called Task Network Modeling (TNM).  TNM allows for the design of the simulation in similar manner to creating a work flow diagram.  The entities on the diagram represent such things as people, patients, documents, or spare parts.  Arrows  are used to connect the entities to show their flow.  Mean times to complete a task are added to provide constraints.  Other conditions are then added to the diagram to determine what sequence tasks are performed and to produce output effects.  The simulator can be used on a stand-alone basis or be connected to a process data source for real-time updates (Barnes, 1996).

Simulation had its roots in equipment development.  The use of MicroSaint expanded this realm into areas that had not previously benefited from the technology.  Simulations can aide in deciding the optimal number of people to use in a given situation.  These situations may be emergency or normal steady-state.  Answers to questions such as “how many people should be on a fire truck or can a pilot handle all situations by himself?” are now possible. MicroSaint helped the military with the development of the Comanche Helicopter.  The Army wished to develop an an attack/reconnaissance helicopter operated solely by the pilot.  The simulation was developed with one key question to answer, “Can one person do it all”?  Four alternative cockpit designs were entered into the simulator.  Variables considered included auditory, visual, cognitive, and psychomotor loadings.  With the variables defined, experiments were performed to capture data on each cockpit design.  The data was fed into the simulator to predict workload over a wide number of conditions.  The final recommendation based on the simulation was that a one person attack/reconnaissance helicopter was not feasible. 

Klein (1998) breaks the development of simulation into two basic types.  In the Monolithic design technique, all relevant information about the model is kept in the model itself.  These systems are usually built from scratch and do not share the components that comprise them.  The decline of this model occurred in the mid 1990’s when the Administration lowered Defense Department budgets.  There became a need for greater efficiency in the creation of the simulation software. The distributed model breaks the total model into a set of sub-models.  These sub-models can be interchanged between systems to lower total development costs.  High Level Architecture was developed by the US Department of Defense and is an upcoming IEE simulation interoperability standard.  Under this standard, HLA defines the behavior of the overall distributed model, called a Federation, and the component pieces, called Federates.  The infrastructure specification details how the Federates communicate through the Runtime Interface (RTI).  Finally, the Object Model Template (OMT) defines the documentation process for Federations and Federates.

(Klein,1998).   Although HLA was developed for military use, Klein provides an example for commercial use.   The example is of a traffic control system. Two methods are examined.  A static development model fixes its elements at the beginning of execution.  The parameters are either hard coded into the model or read in from a data source on start-up.  In contrast, a dynamic model can have its elements change during the actual simulation.   Past development of traffic flow models used Monolithic, static techniques.  Future design will utilize the flexibility of HLA for distributed dynamic simulation. Primary information is used by the core of the model at runtime. It is the variables and the objects that are the basis of the model.  Secondary information is support information for such items as animation that are outside the model’s core. Under HLA, formerly static secondary information can be made dynamic.   In the example above, Federate simulations for cards, pedestrians, and traffic lights are tied together through the HLA Runtime Infrastructure.

University of California San Diego with community clinics used simulation to optimize operations at a local health clinic serving the poor.  The clientele are of little means and the doctors who serve the clinic are more concerned with the welfare of the patients than to make themselves rich.  The clinics are usually run out of donated space in churches or motels.  (Alexopoulos, 2001)  To make each dollar count, the PIP used Industrial Engineering concepts to gather data on the operation.  This data was analyzed and broken into functional relationships.  The Arean simulation language was used to model these relationships.  There are five pieces to the model.  Critical input variables include number of patients in the system, number of providers, average time for each immunization. Design of the interface is made as simple as possible as the staff is not highly computer literate.

(Bulowski, 1997). Scenarios are able to be recreated in a virtual world that would be too unsafe and cost too much to be created in a real environment.  Engineers and architects can use the simulator to determine the speed, direction, and size of fires for any given design.  Evacuation plans can be updated or the building redesigned to provide safer escape routes. The two main components of the simulator are the National Institute of Standards and Technology’s Consolidated Model of Fire and Smoke Transport and Berkeley’s 3-D Architectural Walkthrough software.  CFAST is considered the world’s most accurate fire simulation.  The addition of the Walkthrough software provides an in building view from the perspective of a person having to escape a blaze. As the model becomes more sophisticated, the designers hope to turn it into a training aid for fire departments and building occupants. 

The design of CFAST uses a series of differential equations to move gases and physical properties such as temperature, pressure, etc through the portals into adjoining volumes. •Gas concentration for each type of gas such as oxygen, carbon dioxide, carbon monoxide Volumes are represented by floor to ceiling height and the distance between the walls.  A precise position in space is irrelevant for the calculations.  This is different than the Architectural model which assumes an X-Y-Z coordinate for each room. Similarly, the shape of the vent is entered, but the only other relevant factor that influences the flow of the gases and fire is whether the position of the vent is horizontal or vertical.  

The Walkthru model contains detailed material information about the building area and is a good feeder system to the CFAST model.  Where CFAST contains the chemistry, burn properties, and ignition times, Walkthru provides the detailed geometric and material specifications that architects would need to evaluate strength of building supports and the like. The most important piece of information that Walthru provides to CFAST is the geometry of objects contained in a room. This information is fed into CFAST to provide obstacles to the flow of the fire.  The role of Walkthru is as the visualizer for the fire in progress.  It is designed to show the progress of the fire and smoke and give a visual representation of what can and cannot be seen.  Pictures of how Walkthru views appear can be found at:

S3E2S stands for Specification, Simulation and Synthesis of Embedded Electronic Systems. (Carro, 2000) S3E2S is a CAD design environment to explore the best mix of processor architectures for a desired application. S3E2S minimizes costs by exploring existing processors for an application before the development of new ones.  S3E2S is unique in its ability to model multiple domains with multiprocessor selection.  This allows compatibility over a wider range of hardware.  Time to find a base set of predefined chips is minimized.  The simulator is available to small and medium companies that do not have the resources to develop new chips from scratch for each new product. Besides embedded code requirements, S3E2S also explores the design trade offs of cost, architecture, and power consumption.  It also finds a balance between chip and software based solutions.

(Corsi, 1998) STAR stands for the Simulator for Turbine-Alternator Real-Time Systems.  The power grid simulation provides the user the ability to see system responses under a wide variety of situations such as overload, equipment failure or power dips.  The system provides a visual display of the system being simulated.  Set points and operating variables are fed in by the operator.  The model then processes these inputs and the output responses are then displayed.  These outputs are used as training aids for power plant operators and as aids to engineers.  The engineers use the output response to design for situations not available through live testing. • The real-time simulation unit – a signal computer to provide signal variation • Dynamic Model unit – provides the software to simulate the turbines, grid, etc.

Communications include telephones, public address systems, pocket pagers, and visual pagers. Personnel includes physicians, nurses, technicians. Equipment includes emergency carts or cases with a variety of drugs. Mobility refers to elevators, doors, corridors etc. Detect phase – someone discovers that an emergency exists. Dispatch phase – has two parts. Part 1 – the dispatch system must be alerted (usually by the person who discovers the emergency). Part 2- the system must alert the team members and direct them to the site of the emergency. Deploy phase – specific medical personnel must travel to the patient treatment location, special supplies must be obtained and transported and certain facilities must be prepared. Deliver phase – the time period during which the members of the emergency team are treating the patient. Disperse phase – occurs when the emergency is over and represents the time required to restore the initial settings (return material to its standby location) The program has been done in a modular manner in order to permit the evaluation of a wide variety of hospital configurations and emergency mobilization approaches The system has been tested over a period of two years (1966, 1967) and 229 cardiopulmonary arrests. Significant improvements have been reported. The system was tested on an IBM360-75 machine.

Operation panel was realized by a portable PC and operated by the crews according to the captain’s orders. It fulfils data installation, working state choice, voyage control etc. Display panel is realized by another portable PC and backups with Operation Panel mutually. The interface cabinet consists of intelligent I/O board, high speed RAM and power modules. It is designed to fulfil the real time communication with the rest of the systems. Motion modeling – underneath water model, near surface model, submergence and emergence maneuvering model, emergency maneuvering model Software modules – submarine motion module, navigation devices simulation module, sea environment module, information communication module, image display module, module of PID controller or rudder Simulation can not only emulate the motion characteristic of the specific submarine, but also train the skills of crews in emergencies. Video and audio effects can provide VR circumstances for the crew. The simulator can be used as a drill device for training and also as a method to study the maneuvering performance of a submarine.

Harpoon missiles used by S-3B aircraft – a twin jet antisubmarine warfare carrier based aircraft, carrying a crew of four: pilot, copilot, tactical coordinator and sensor operator. Built by Lockheed California Company HACLCS provides the means for preparing and launching the missiles. It includes Harpoon Aircraft Command and Launch Control (HACLC), two Control Distribution Boxes (CDB), two Fire detection Control Units (FDCU) and two missile umbilical cables. Computer mode is highly automated using the capabilities of both the aircraft computer and the HACLCS. LOS provides a backup launch capability in the event of degraded equipment performance

Abnormal mode includes hung missile, digital fault or missile fire. Normal indicates a missile ready to fire. The various capabilities are to be simulated by the system and to replicate exactly the real-life behavior of the Harpoon missile.  The operator has control of such parameters as mission target, time to launch, arming, and launch. 

ASIA – simple agent-based simulation software, implemented in Java. Applications for economic and environmental studies including the international greenhouse gas (GHG) emission trading. Social layer – describes the basic role of agents in the society. Central, Participant and Watcher agents were implemented. Central agents create, register and initiate Participant and Watcher agents. A prototype for the greenhouse gas emission trading under the Kyoto protocol was developed. Gaming simulations with human players in an environment similar to the agents’ environment are expected to help in constructing plausible behavior models and extracting essential dynamics. The gaming simulations are planned to be executed at several universities.

ACT is a network of powered horizontal and vertical monorail conveyors which permits the movement of materials and supplies from the centralized supply processing and distribution center to the various decentralized areas throughout the hospital and vice versa Factors that need to be studied include; the mix and numbers of carts – there are nine types of carts that are to be employed, each type having a different purpose; number of transporters; manpower – there will be a significant number of required personnel that will handle the system. There will also be interactions between people from other departments and the system (such as people from the food service department, accessing the system to transport food). An optimal schedule would have to satisfy all the delivery demands: minimize capital investment, provide staffing patterns for every unit, minimize travel time etc.

Concept 2: Use this list in order to monitor the flow of carts in the simulation. This requires a priori knowledge of the precedence relationships for each event within the system. WIDES is a Fortran and GASP based event simulation language with event-oriented perspective. It is composed of more than fifty subroutines, used as model building blocks in the process of modeling the ACT system Database - contained the input of the system – user demands for materials (scheduled or unscheduled) Simulation program – 2 modules:  the initiation module (initialize system variables and parameters) and the execution module. Output – messages are constantly output on the CRT. Three types of hard copy output: WIDES trace, ACT movement table and ACT transportation table With what has been demonstrated through simulation, an erroneous operational plan has been averted from being implemented in the new facility. The simulation has proven to be an invaluable educational tool for all those involved with the ACT system.

Communication is key to successful operations of business, military and health care systems. In general, two types of communications are sought: dynamic data exchange during simulation runs directly between simulations and sharing data between simulation runs through a central repository. CATT – a technical approach for providing mission training for helicopter pilots on the dynamic conditions and alternative courses of action. The user would be able to access a GUI component that would communicate through COM Services with the Micro Saint simulations. CART – developed under the Air Force Research Laboratory. The addition of an adaptive simulation interoperability environment allowed CART models to communicate with other simulations.

GPS greatly eases the task of building computer models for certain types of discrete-event simulations. It lends itself particularly to modeling systems in which discrete units of traffic compete for scarce resources. GPSS has been applied to the modeling of manufacturing systems, communication systems, computer systems, transportation systems, health care systems. It has also been used in chemical engineering, mining engineering, and cancer research. Only seven statements are required to model a one-line, one-server queuing system in GPSS. Beginners can learn quickly how to use it. Can be used on a variety of systems: MicroVAX I, MicroVAX 2, MicroVAX 3, Apollo, Optimum 5/10, Optimum V, IRIS, IRIS Turbo, Sun 3, IBM PC etc. GPSS-FORTRAN, APL-GPSS PL/1 GPSS – all these languages have GPSS embedded.

Recent implementations have GPSS embedded, so there is no need to call HELP routines. GPSS is trivial to learn only to a superficial depth. Real mastery of GPSS requires considerable study. Misconceptions about the lack of power of GPSS come from people with an insufficient grasp of the language. Early versions were batch oriented. Current versions are designed for both interactive and batch use.

Two broad classes of simulation practitioners – simulation software developers (research model) and problem solvers who use simulation (practice model) Both medical education and simulation education require the student to perform certain tasks, after acquiring an extensive knowledge (the patient in the case of the doctor, or the system to be modeled and simulated in the case of the engineer). Basic coursework – foundations of decision modeling. These courses should focus on applied problem solving using the models rather than derivations. Additional coursework – all students should have enough coverage to feel comfortable dealing with non-technical management in businesses and other organizations

Service processing requirements has always been a difficult problem to solve and it is of overwhelming importance in certain service sectors such as hospitals or the military. Queuing theory has emerged as a useful decision making technique. Often times, the use of simulation modeling has been required in problem analysis. Order requests arrive in batches in a random fashion with an average of 1.4584 per day. The total service process is comprised of several service activities that must be performed in sequence. Average processing time for pre-USPGO orders is 45.4 minutes and for post-USPFO orders is 80 minutes. A single line forms for all orders waiting for processing. The sequence in which orders are served is first-come, first-served. Five models were developed. The most successful was model 5, which had introduced additional clerks when the number of orders in the queue reached a particular level. Manipulation of the queue size and clerk number was effective in reducing overall processing time of the orders

Automatic Support Detachment (ASD) provides the development and support of all telecommunication systems for the HQ Department of the Army. The current system was designed to handle 1500 messages per hour with an average message size of 2000 characters, operating on a round the clock basis. The proposed improved system would handle 6000 messages per hour. Three models  were developed, representing the Front End Control, Message Processing and Staff Service Center Subsystems. Applications would fall in eight categories: control programs, I/O, video display, message processing, table subsystem, data management subsystem, offline programs, COM subsystem. The language could model multi-CPU system, is easily modifiable, facilitates the linking of several modules to form one conglomerate system, allows tracing of all of the simulation activities. The objectives of the modeling effort were to determine effects of hardware conversion, system design and software modification. A flexible interface between the data collection effort and the PCTCS model was designed. User-written simulation reports provide a clear picture of the “system life’ of the simulated messages. PCTCS model will be used to evaluate design alternatives and proposed system enhancements.

Regular users generally don’t like change. When simulation is employed to solve problems, users are generally management or industrial engineers, not the providers to whom the solutions are routinely applied. Without a comprehensive introduction to the tool and its potential, non users view simulation based recommendations as ‘black-box” answers to complex problems. To gain acceptance of any solution, regardless of its source, it is imperative that every member of the affected group be involved in the decision-making process. The more complex the problem, the more critical management support and commitment become. There is considerable resistance to the dehumanizing nature of time and motion analysis. Most workers in general, and health care providers in particular, regard such things as treatment-time evaluation and standardization as unreasonable and unrealistic. Healthcare environment is far more complex than any industrial environment. The idea is to be as flexible as possible and to avoid being imprisoned by your own methods. Take advantage of all the opportunities offered by opposing views by evaluating the model’s sensitivity to its inputs.

Simulation models emphasize the direct representation of the structure and logic of a system as opposed to abstracting the system into a mathematical form. The availability of system descriptions and data influences the choice of simulation model parameters, as well as which objects and which of their attributes can be included in the model. Alternatives can be assessed without the fear that negative consequences will damage day-to-day operations as would be the case if experiments were conducted directly on existing, operating systems. Variation has to do with the reality that no system does the same activity in exactly the same way or in the same amount of time always. Variation may be represented by the second central moment of a statistical distribution, the variance. Variation may also arise from decision rules that change processing procedures based on what a system is currently doing, or because of the characteristics of the patient receiving care.

Public policy applications have to do with evaluating strategies for delivering health care that are implemented in state or national policies. Example – a simulation model for projecting the number of physicians, nurse practitioners and physician’s assistants in Indiana, from 1975 to 2000 as well as the demand for primary health care. Example – a family therapy process called Brief Systems Family Therapy (BSFT). Six checkpoints in this process were specified; problem formation, solution formation, enactment formation, hypothesis formation, task intervention formation and follow-up. Determining the level of capital expenditure for equipment needed to effectively provide patient care is an important application area for simulation in health care delivery. Example – a simulator for helping to establish the resource requirements of a small animal veterinary practice There are many applications of simulation to operational policies of health care providers. Example – an application of simulation to emergency room operating procedures. At issue was the excessive length of time non-urgent patients waited for care in an emergency room at a non-profit hospital.

Internet based simulation is difficult to learn and too easy to make mistakes which have disastrous consequences.  Mistakes such as faulty use of pointers are difficult to find. There is no inherent mechanism to describe parallelism; Internet debugging tools are simulation unaware, i.e. they operate at a level far below that which would be convenient and necessary for most simulations. HLA is a simulation interoperability standard currently being developed by the US Department of Defense. The architecture is defined by: rules which govern the behavior of a distributed simulation (federation) and the individual distributed components (federates); an interface specification which defined the interface between each federate and the Runtime Infrastructure (RTI); an Object Model Template (OMT) which provides the framework for defining federations and federates. HLA defines a two part interface which federates are required to use for communicating with the RTI. It is based on the ambassador paradigm. A federate communicates with the RTI using its RTI ambassador. Conversely, the RTI communicates with a federate via the federate’s ambassador. From a programmer’s point of view, the ambassadors are objects and the communication is done by calling methods of these objects. Being a member of a distributed simulation imposes some general problems that the stand-alone simulations do not have to deal with: synchronization, data exchange, and data representation.

This is the most obvious solution. If the source code of a tool is available and well documented, this is the most straightforward and probably the least complicated solution. Some simulation tools translate model descriptions written in a tool-dependent modeling language into another programming language. This intermediate code is then compiled to an executable file. It is possible to modify this code to realize the HLA extensions. This solution is well suited for tools that offer an open and extensible architecture. The tool should offer a library interface with the ability to call functions or methods in these libraries. Additionally, the tool should make it possible to implement callback functions or methods. The last solution for tools which can not be connected to the RTI by any of the prior methods is the development of a gateway program. The gateway program could communicate with the simulation tool via appropriate means (files, pipes, network) depending on the capabilities of the simulation tool.

Verifying the model is a process of comparing the actual patient flow with the on-screen patient flow. The various documents (flow charts, arrival rates, etc) should be combined with records of the various team meetings to form an “assumption document”. Validation involves testing the model to ensure that the actual system length of stay times are mirrored by the simulation model. This particular case model was particularly difficult to validate. Any patient type with significant admission rate would not validate. Final three stages are related to setting up alternatives, run each alternative, evaluate each alternative and choose the one(s) that best suit the initial goal. Successful simulation studies are dependent on the cooperation of each department that is affected by the study and that affects the objective of the study.

Simulation models in Java can be made widely available. An applet can be retrieved and run and does not have to be ported to a different platform or even recompiled and linked. Java provides a high degree of dynamism (applets run in browsers). Java threads make it easier to implement  simulation following the process interactive paradigm. Built-in support for animation. Simjava – a process based discrete event simulation package for building working models of complex systems. Facilities for representing simulation objects as animated icons on screen. DEVSJAVA – an environment based on DEVS (Discrete Event System Specification). It supports High Level Architecture, agent based modeling and System Entity Structure JSIM – an environment supporting web based simulation as well as component-based technology. The environment uses Java Beans to create reusable components. Simulation can be built either using the event package or the process package. JavaSim – a Java implementation of the original C++SIM Simulation toolkit at University of Newcastle upon Tyne, UK. Silk – commercially available general-purpose simulation language based around a process-interaction approach and implemented in Java. WSE – Web-enabled Simulation Environment) combines web technology with the use of Java and CORBA.

Major market trends are driving manufacturing from mass production to mass customization. Manufacturing enterprises could follow a virtual manufacturing operation composed of three modules: an agent architecture that decomposes the system to address both information modularity and the physical realities of manufacturing; a simulator and an infrastructure to support the implementation of the agents. Exploration, discovery-based and learning by doing are valuable methods of learning which give a learner the feeling of involvement. Learning how to build models is best done by actually building models. Various military training applications are the bulk of training environments that are migrating to the web. There was a lot of use of simulation in the military domain, as it could be seen in the previous slides. A web-based system called ASTAR (The Army Standards Repository System) was developed by the Army Model and Simulation Office to enhance the army’s decentralized, consensus based standards development process. A web based tool facilitates the Standards Development Process. Large scale computer simulations can take days to run and produce massive amounts of output. An example of a scientific simulation application is the Weather Scenario Generator – intended to mine a very large array of environmental data and provide results to a user at interactive speed. Web based simulation of autonomous software agents is another example of web based environments used to explore the potential of the web and new software technologies. Another example is the problem of controlling crowds in public places. Simulations can be used to model these problems.

(Hubal, 2003)  Federal and corporate funding were used to develop a training aid for police officers who need to deal with the mentally ill.   The application is called JUST-TALK.  It is built upon five basic scenarios.  All scenarios involve a young, adult male behaving strangely.  The youth walks into the middle of a street and is almost run over by a car.  The officer responds to the report and finds the youth either sitting on a bench, standing near it, or pacing around it.  Through gestures and a natural language interpreter, JUST-TALK allows the officer in training to determine if the virtual human is schizophrenic, paranoid, depressed, sad, but normal, or stressed but normal.  The model is based on the interaction with a lifelike virtual human which displays emotion and responds to commands in some manner.  The developers of the software believe that the simulation along with classroom instruction and role playing will lead to an effective learning environment.

(Hubal, 2003)   JUST-TALK utilizes an architecture called AVATALK to allow conversations between people and virtual humans.   The human user then gets to see and hear responses from their virtual human subject.  The system is broken into three main components: 1) Language Processor – this module breaks down human speech into a semantic representation suitable for interpretation.  Once interpretation is performed, the response is fed back from the behavior engine.  The Language processor works in reverse and speaks or formulates a facial or hand gesture.  2) Behavior Engine – dynamically loads the context and the knowledge needed by the Language Processor.  3) Visualization Engine – takes gesture, movement, and speech output and uses a 3-D virtual human to perform the requested actions.  The mouth moves to lip synch the words using a morphing of the 3D model and playing selected animation frames created by motion capture techniques. 

Situation awareness is hard to keep up for fighter pilots while outside of live action zones.   The simulator allows skill development without the expense and risk of live flight.  While nothing can replace live flight, simulation is a key training tool.  Simulation is high on the list of requested activities by pilots, training managers, and air weapons controllers.  McDonnell developed the SIMNET program standing for “Simulator networking” under sponsorship of the Advanced Research Projects Agency and the Army.  The key feature of the simulator was the use of independent processors in a distributed environment.  This allowed independent update of target and pursuit participants.  The design utilized F-15 cockpits with stick and throttle controls.  Each has a color touch screen front panel. 

The simulator was rated higher than actual flight training in a number of different areas.  The reason for this satisfaction is clear.  At no time is the pilot actually at risk.  In multi-bogey (enemy airship) situations, the skies become crowed and the risk of a mid-air collision increases.  Similar reasons exist for the other high satisfaction areas.  For number 2, no missiles are actually being fired at the pilot.  The evasive maneuvers can be practiced without risk. Electronic countermeasures can be employed without loss of radio contact or radar.  Visibility can be changed during any heavy weather.  Low altitude can be practiced without crashing into the ground. Using touch screens instead of rudder pedals did not lower the pilot’s opinion of the simulator.

Despite the fact that the pilots were highly satisfied with the design of the simulator in many areas, some areas were lower in satisfaction than real flight. The images that were projected on the forty foot dome were not as sharp and clear as the view in the sky.  Formation variations and realism was limited as the projections were at a fixed distance away from the cockpit.  Mutual support to attack bogies was not as realistic as combat situations.  The biggest influence on the overall satisfaction was the software version of the radar.  The simulator used an older version than the one in the F-15.  This changed tactics on occasion.

(Cavazza, 2003) Virtual humans have become important in the development of intelligent user interfaces.  Their role can be as intelligent assistant and instructors or become part of the actual training scenario as a patient. This simulator integrates a knowledge-based system with a 3-D virtual patient.  The medical students use the system to learn decision making during cardiac emergencies. The work was spun off from Badler’s work on battlefield casualties.  In Badler’s system, field medics would learn how to deal with emergency treatment.  Cavazza’s simulation differs by using a physiological model of the patient’s internals.  Prior to invasive action, the trainee can also see the patient’s appearance and read-outs that would appear on a hospital’s heart monitor. 

(Cavazza, 2003) The appearance of the patient varies according to his or her vital signs during the coronary emergency.  The input to this appearance includes the patient’s Heart Rate (HR) and Mean Arterial Pressure (MAP).  Complimentary examinations reveal other parameters.  For example, if a cardiac catheterization is performed, the trainee can obtain Pulmonary capillary pressure (Pcap). Instructions to the virtual nurses are performed via speech recognition using a simple command language.  This language is comprised of a simple command language and its parameters.  For example, a command might be “Inject drug {Drug name, dosage}”. An initial state of clinical conditions is given to begin the event.  These conditions propagate to a reaction set of symptoms requiring diagnosis.  When a treatment is given, the symptoms are updated until a steady state is reached.

The virtual presentation of the patient and the surrounding area are representative of the intensity and gravity of the ER situation.  The lighting, available equipment, patient’s appearance, background noise, and in room sounds intensify based on the condition of the patient.  This allows the realism needed to simulate a life and death situation. As needed the trainee can turn the patient and view the equipment placed in the room.  Monitors can be viewed to show the patient’s vital signs.  Nurses respond to voice commands or to those given through a series of menus.  Drugs can be administered.  Animation of the patient shows response to the various treatment given.  This animation might show labored breathing or writhing.  All aspects of the 3-D environment are used to provide as much realism as possible.

(Duncombe, 1997)  Mobil Subscriber Equipment (MSE) is used by the Army for battlefield voice and data communications.  The System Control Center (SCC) provides battlefield personnel with orders and looks to receive reports back from the field.  The SCC is to be replaced by the newer Network Mangement Tool (NMT).  To prepare the troops for the new NMT, GTE corporation created the Communications Network Simulator (CNS).    This simulator provides tactical situations requiring  a precise execution of tasks.  The first area that operators are trained in is the setting up of the system.  These processes include the software and communications network initialization and to connect the system to the message generator.  The operator also learns to deal with normal operations, degraded operations, maintenance and administration of the system and shutting and tearing the system down. Messages flow from the NMT at a simulated rate of 231 per hour based on a plan.  At least 130 of these messages can be from special communications outside of the active area. These special areas include Line of Sight (LOS) radios and NATO interfaces.  Emergency situations simulated include loss of partial communications, loss of equipment, and disabling of personnel.

ISD was one of the first systematic training models ever developed. It was used in World War II by the US military to train soldiers in aircraft recognition. This is a synthesized high-level framework for evaluating training effectiveness. The approach is based on the ISD process and contains six major components: training task identification, training proficiency evaluation, training task prioritization, identification of simulation training support, simulation training execution and feedback. The methodology requires tasks be identified in advance of training. Each task is then ranked, relative to others, by the military unit undergoing training. The weights reflect the importance of each task to mission accomplishment. Precise measurable standards are best for evaluating the training proficiency of the military unit.

Strategic directions are areas in which simulation can be applied immediately, but where we have not taken full advantage of the technology that is available. It is possible to embed simulation modes in the operating systems of computer systems. These systems can feed data about their operations into a data store accessible to the simulation process. Periodic execution of these would evaluate the performance data and identify operational trends. The world is filled with opportunities to apply computer simulations to assist in real-time decision making. Any place that information is available in a digital form and humans are evaluating that information to make decisions based on it, there is an opportunity to support the human with simulation. Combat consultant, aircraft navigation or crowd management are just three examples of the potential areas of application. We need to create virtual environments that are persistent over many years and that form the foundation for specific studies, training and entertainment that will be conducted within them. The gaming community has already done some steps in this direction. Similar virtual worlds need to be created by high level sponsors of studies and training events. In the simulation business we strive to create virtual worlds that accurately represent the real world. However, there are few simulations that portray a really convincing virtual world. Statistically accurate simulations are excellent for many applications, but we need to focus also on the richness of the environments.

Many simulations are driven by statistical distributions that characterize the average behavior of a system, but do not claim accuracy for individual events or small time intervals. These distributions do not model instantaneous behaviors of intelligent or reactive beings. We need techniques for inserting intelligent, reactive, unique human behavior in the virtual world. It should be possible to develop an architecture that supports an entire domain of simulation systems, providing a large common pool of functionality. Distributed simulations cannot exist without sufficient reliable communication bandwidths for delivering events and synchronizing execution of the entire system. This bandwidth is currently one of the limiting factors on the size of a distributed simulation. Since bandwidth is a problem for every internet based application, a lot of commercial research is being done in this direction. A lot of research has been done to discover techniques for practical and efficient synchronization of distributed simulation processes. We must identify applications that are well served by the different modes of event management. Research needs to find a practical and valuable home in commercial, government and military simulation systems.