A-TEAM:

Advanced Training System
for Emergency Management

2. OBJECTIVES

The main objective of the A-TEAM project is to improve the learning process in complex, technical domains, using the example of technological emergency management. Improved learning is achieved by integrating information technology (dynamic simulation, visualisation, GIS, expert systems and case-based reasoning) within an innovative didactic framework that fully exploits the potential of multi-media information systems. Improved efficiency and effectiveness of the learning and teaching process in complex technical domains difficult to cover with traditional didactic methods is seen as an important contribution toward a technological and information society.

The project will develop and test a new approach to advanced technical training using an integration of artificial intelligence (AI) technologies and dynamic simulation modeling to create fully interactive multi-media content within a real-time knowledge-based system framework for the domain of emergency management applications. The underlying client- server architecture supports easy access in Intranet/Internet distributed systems.

The training for the management of technological and environmental emergencies faces the difficulty of a domain where realistic experimentation is impossible and personal experience is scarce. Learning-by-doing, the by far most effective method for practical skills in a very complex and ill-structured domain with an enormous range of possible situations, is restricted to rather limited and somewhat artificial exercises, which also carry substantial costs. The possibilities of computer simulation combined with artificial intelligence and virtual reality offer an interesting new approach to this dilemma.

A new paradigm for computer-based learning:

This integration of AI methods and dynamic simulation supports the real-time creation of multi-media content for a radically new paradigm in computer-based learning systems with realistic interactive training scenarios instead of traditional textbook-derived content. It introduces a learning-by-doing concept in a virtual reality mode for an application domain where practical experience is necessarily scarce. Didactic elements: Integrating additional didactic elements such as explanatory material in hypermedia formats, trace-back and log functions, multiple choice tests and evaluations, and a performance based adaptive structure of training units, will allow to use the system for both individual and group training across a broad spectrum of technical competence.

The system guides the learner through a simulated emergency, and monitors her reactions to the systems requests for information or decisions. Any deviation from the expected optimal path will lead to an intervention of the system. The error will be pointed out, in-depth background information on the concepts involved will be provided. Learner comprehension will be verified through embedded multiple-choice tests or simple training exercises, until the overall training case can be resumed. By logging and time-stamping any learner response, the system can evaluate the training run in a post-mortem analysis and provide constructive critique including suggestions for further study.

Flexible didactic framework:

To provide a flexible didactic framework, the system is embedded in a hybrid forward-backward chaining real-time knowledge-based system (KBS). The KBS guides through the simulated emergency just as in a real emergency management application and co-ordinated the information resources, including the simulation models. At the same time, it monitors the trainee's responses, and can trigger additional explanatory material, tests and questionnaires, or modify the sequence of events such as returning to a previous stage to re-run a critical part.

Case-based reasoning training scenarios:

Case-based reasoning within the framework of the overall real-time KBS will adapt these methods to the specific training scenarios. The possibility to move between levels of abstraction, i.e., go from a semi-empirical method to the underlying physical phenomenon, or abstract from a specific simulation to a generic class of cases interactively, adds an important experimental and explanatory component for the learning process.

Fully interactive immersive experience: The project aims at providing fully interactive immersive experience through dynamic 3-D modeling of emergencies and their impacts with high- performance visualisation. This allows the trainee individually or in a team to gain experience in a domain where experimentation is largely impossible and the real-world experience necessarily very limited: low-probability high-consequence events such as chemical spills, fires, and explosions related to the chemical process industry and hazardous goods transportation. Integrated advanced simulation models: A major objective is to integrate advanced simulation models such as 3-D CFD codes and probabilistic models for the simulation of accidents and their consequences. This will provide realistic training scenarios, e.g., the plausible accidents defined in the Seveso II directive (96/82/EC); the use of dynamic 3-D graphics and dynamic GIS will support an intuitive understanding of complex physical and chemical processes and phenomena and make the training situations intellectually challenging.

True interaction:

A related objective is to make the system truly interactive despite the heavy computational demand of simulation based multi-media content and the corresponding high- performance visualisation required. To guarantee the fast response that a fully interactive real- time training system requires, both to capture the uninterrupted attention of the trainee and to convey the real-time nature of emergency management, the detailed simulation models will be implemented in a parallel (cluster) computing environment; they will also be augmented by the simplified methods more commonly used in daily practice, but based on machine learning (neural networks, ID3,...). The training cases for the machine learning will again be generated by Monte Carlo simulation.

Technical components:

From a technical point of view, the system will integrate:

  • a real-time hybrid forward-backward chaining knowledge-based system as the driving engine of a training session;
  • embedded 3-D dynamic simulation models and high-performance visualisation and animation of the model results;
  • an XML/CSS based structure of the didactic hypermedia material;
  • VRML and Java3D interface components for the Internet access;
  • student management and evaluation features.

Multi-media contents:

From a content point of view, the system will be based on the experience of several domain experts, trainers, and end users involved in the project. The contents will be represented in the systems knowledge base, the models and their examples case library, and the accompanying didactic hypermedia material as well as background information such as a set of Material Safety Data Sheets (MSDS) for hazardous substances, including, inter alia, first aid instructions.

Multiple languages:

Another objective of the project is to build A-TEAM as a multi-lingual system and develop efficient approaches for managing multi-language content. Taking advantage of the project consortium composition, the system will be available in several languages (English, German, Italian, French, Spanish, Portuguese and Greek). Co-ordinated multi-lingual development will be one of the technical issues addressed.

A range of test cases:

To help meet the objective of improved the functionality, usability and acceptability of this information product and related services, the A-TEAM training system will be tested in a range of class-room and Internet access set-ups, with different target groups from fire fighters to safety engineers from public and corporate environments.


© Copyright 1995-2016 by:   ESS   Environmental Software and Services GmbH AUSTRIA | print page