[an error occurred while processing this directive]

Weather Forecasting Group

WEATHER FORECASTING HOME WEATHER FORECASTING STAFF WEATHER FORECASTING EXPERIMENTAL RESULTS WEATHER FORECASTING COLLABORATIONS WEATHER FORECASTING ANNUAL REPORT

Intelligent Alerts ARC Linkage Grant
(RMIT/AOS/BoM)

Open Agent Architectures for Intelligent Distributed Decision Making

Grant summary

Sophisticated software systems are part of the essential infrastructure of our daily lives. Complex systems such as the internet, finance, or telecommunications software cannot have a centralised point of control or a single developer. The aim of this project is to develop an architecture and support infrastructure enabling intelligent agents to locate and use services in such open systems. The fundamental questions that must be addressed include issues such as how agents will find and use newly added services and how services will communicate with each other, given that they are developed independently.

See grant # LP0347025 on the Australian Research Council website for this grant.

Partners

People

Links

Initial Project Proposal

From meeting of BoM/RMIT/AOS ARC project, 22 May 03.
Present: Sandy, Lin, Mal, Nick, Ralph, Ian, Tom Keenan, Michael.

Project name: (tentatively) Agent Based Bureau Alerts (ABBA), or Weather on the Way (WOW), or Jack Agents For Forecaster AlertS (JAFFAS)

Concept: Broadly, a system with agents attached (1-1) to aircraft flights across Australia (or Pacific) that discover agents attached to individual weather events en route, and react or issue alert if a change in flight planning is required.

More detail: ingest from airline or Air Services Australia flight plans (in whatever format they are available). Spawn new agent for each flight. Similarly, for weather events, ie Titan thunderstorm detections, spawn an agent. This may involve forecast thunderstorms, or future forecast track of an existing thunderstorm together with probabilities. Agents would communicate potentially many-many, ie we are not talking client-server, but peer-peer.

The flight agent will need to "discover" all relevant weather event agents along its path, or future contingent paths, assess their impact, and issue an alert to the airline flight planning software or its operators. Does flight discover weather, or does weather discover flight? Could go either way. Or do we use waypoint (regional) agents for each to register with (same model used by air traffic controllers) to handle discovery (but not actual relevance /conflict detection)?

'Discovery' will probably involve a broker (or network of brokers) that agents subscribe to to be informed of events of various types. Discovery issues are:

  1. New status for known event.
  2. New event of known type.
  3. New (unknown) event/service type (eg. volcanic ash).

Weather events could be: thunderstorms, fog, changes in TAFs, discrepancies from TAFs in AWS data.

Alerts could include the removal of a previously threatening thunderstorm (anything that could effect the flight plan), or changes in the probabilities of events along the flight path.

References and other documents

Sandy Dance, 25 July 2003
[an error occurred while processing this directive]