Objective Output Schedule |
|
The DiveArch project is collaboration between Susanne Jul (NRC Research Associate) and the Pacific Disaster Center (PDC), funded jointly by the National Research Council and the PDC.
The goal of the DiveArch project is to design a computational architecture for interactive agent-based Disaster Information and Visualization Environments (DIVEs), and to prototype an application development framework implementing that architecture. The DiveArch framework will build on an existing general-purpose agent-based platform, adding an ontology of disaster-related agents and shared runtime conceptualizations of disaster-related system parameters (e.g., regional topography or time of day). It will incorporate an existing GIS package for managing geospatial information, as well as a set of visualization and interaction techniques that are appropriate to disaster-related information. Evaluation of existing general-purpose frameworks, and selection of agent and GIS frameworks that appear best suited to the needs of the disaster management community, in general, and the PDC, in particular, is an important part of the project.
Concurrently with architectural design and framework development, the DiveArch project will work with two application development projects. These projects will serve as case studies to ensure that DiveArch identifies and addresses the needs of DIVE development accurately. Case study projects are expected to be selected from among current PDC and external efforts, and will benefit by the additional architectural and, to some extent, programming support.
Vevac: A toy agent-based model of vertical evacuation. Interactive demo developed in NetLogo.
Jul, S. (2007). Who's Really on First? A Domain-Level User, Task and Context Analysis for Response Technology. Proceedings of the 4th International ISCRAM Conference. Delft, the Netherlands, May 2007.
Comprehensive list of potential criteria for selecting an agent-based modeling platform. The list is intended to be used with a satisificing selection strategy, not for comprehensive comparative evaluation, i.e., to select high-priority criteria that can then be used to limit the candidate pool.
Agent-Based Modeling, the PDC and DiveArch. PowerPoint presentation.
Last updated: 24-Apr-2007