Bushfire emergency response simulation

Abstract

Improving the ability to model, manage and communicate the impact of uncertainty in complex interacting systems is becoming increasingly important. This is particularly important for defence as the nature of warfare is rapidly changing due to the incursion of novel defence technologies. The dominant attribute of modern warfare is its network-centric characteristic, which can be both an enabling and a limiting factor. Network-centric warfare refers to a potentially large number of entities or assets with varying capabilities for autonomous decision making whose coordinated fighting power is enabled through information exchange within a constrained system structure that can be described in terms of interacting networks. Such networks are known to be able to exhibit complex behaviours such as rapid phase transitions, oscillations, chaos, disturbance rejection, and adaptation. Understanding the potential for complex behaviours associated with particular military operations is essential to establish appropriate simulation and data collection processes that inform decisions about force design and asset acquisition. To explore these concepts in an unclassified manner, this project developed a bushfire emergency response simulator as a surrogate for similar Defence problems. This paper describes our multi-agent simulation design which is intended to capture key dynamic characteristics of emergency response operations including individual fire response asset behaviours and team hierarchies in perception and decision making. This new bushfire emergency response simulator includes assets to be protected, command and control assets, fixed and mobile sensing assets, and assets within the response team with perception, decision, action and communication capabilities and with different capabilities to move and fight the fire. A new cellular automata type bushfire model with dynamical cell interactions considering fuel and wind is introduced, that enables simulation of a greater range of fire behaviours with higher fidelity, than simple cellular automata models.This is the second in a series of papers describing bushfire emergency response simulation as a surrogate for Defence problems to explore the quantification of uncertainty in modelling, simulation and analysis of complex systems. This paper describes in further detail the bushfire emergency response simulator which was applied to uncertainty quantification in Bruggemann et al. (2019).