Cloud computing based bushfire prediction for cyber–physical emergency applications

Highlights

• A novel cloud based framework to deploy/process fire models within a deadline.

• A novel scheduling mechanism integrating user’s req. and minimising resource usage.

• A case study using Tasmania Bushfire Model for evaluating the Cloud based framework.

Abstract

In the past few years, several studies proposed to reduce the impact of bushfires by mapping their occurrences and spread. Most of these prediction/mapping tools and models were designed to run either on a single local machine or a High performance cluster, neither of which can scale with users’ needs. The process of installing these tools and models their configuration can itself be a tedious and time consuming process. Thus making them, not suitable for time constraint cyber–physical emergency systems. In this research, to improve the efficiency of the fire prediction process and make this service available to several users in a scalable and cost-effective manner, we propose a scalable Cloud based bushfire prediction framework, which allows forecasting of the probability of fire occurrences in different regions of interest. The framework automates the process of selecting particular bushfire models for specific regions and scheduling users’ requests within their specified deadlines. The evaluation results show that our Cloud based bushfire prediction system can scale resources and meet user requirements.

Introduction

Due to human activities and climate changes, bushfires have increased dramatically in the last few years  [1], [2]. Every year thousands of acres of forest area is destroyed that includes not only loss of several animal and plant species but also human lives and properties. For example, during the Black Saturday 2009 fire, one of the most significant disasters in Australian history, 173 people lost their lives and 2298 homes were destroyed along with several other environmental losses. Therefore, forest fires are considered to have serious environmental and socioeconomic effects that are aggravated due to increase in climatic temperatures.

In response to this, several fire prediction and behaviour models have been developed during the last four decades to reduce the after-effects of bushfires. Several desktop based fire simulation tools are available that incorporate such models. Some well known tools are SiroFire simulator  [3], BehavePlus  [4], FARSITE  [5], Spark  [6] and HFire  [7].

In general, the estimation of fire risk and fire spread are dependent on several geospatial input data sources, some of which are dynamic and change with time. For example, weather data changes with time and space. Furthermore, each user may want to do computation for a different geographic extent and at different spatial resolutions which defines the amount of input data, storage and computational resources required. Due to the complexity of computation involving data of different formats, sizes and from different sources, the data processing is not a trivial task and may involve expensive investment in terms of computational hardware, software and deep computing skills. Furthermore, although most of these simulators help us to understand in an efficient way and in an accurate form, it is still quite manual and time consuming from the perspective of a user who has little knowledge about underlying infrastructure.

Some of these drawbacks were addressed in fire management systems such as Virtual Fire [8] which allows an easy to use web interface to access and visualise different data sets including on-demand fire behaviour simulations. Most of these fire prediction tools and technologies are designed to either work on single desktop machines, clusters or limited high performance computing. Thus, these systems suffer from low scalability and availability  [9].

Recently, several researchers have begun to see Cloud computing technology as a cost-effective and highly scalable solution to Big Data problems in different domains such as geospatial sciences and threat management  [10]. Cloud computing provides elastic and on-demand access to an almost infinite amount of storage, network and computational resources  [11]. Due to the pay-as-you-go model of Cloud computing resources, users do not have to maintain expensive computing facilities or face up-front cost. Thus, Cloud computing infrastructure allows elastic storage and computational capabilities for managing a fluctuating number of user requests. Some researchers have already showed the benefits of Cloud computing which provides dynamic and scalable computing and storage infrastructure  [12], [13].

Despite so many benefits offered by Cloud computing, the solutions available for tackling real geo-spatial science problems are limited. Some studies used Cloud computing for storing and managing a large amount of geo-spatial data but using their infrastructure with a strong manual component  [14]. Others only used Cloud computing to increase computing capacity  [15], [16]. Most of this work does not offer an effective solution as it neglects either user requirements (e.g. deadline) or still has a large manual component. During emergency situations such as bushfires, even a small delay can result in the loss of many lives. Thus, making these solution unpractical for time constraint cyber–physical systems [17].

Over the last several decades, there have been several deadline based scheduling algorithms for scheduling applications in a Cloud computing environment  [18], [19]. As they are developed for specific application domains, they cannot be applied directly to scheduling of bushfire prediction application.

To overcome the limitations of previous bushfire prediction systems, we propose a Cloud based fire prediction service framework that not only allows access for multiple users simultaneously but also considers the requirements of each individual user. The proposed service also minimises the cost by keeping Cloud resource usage to a minimum. The proposed framework also allows users to use different bushfire models according to their area of interest. We also evaluated the proposed framework using a bushfire case study from Tasmania, Australia. In summary, the main contributions of this work are:

• A novel architectural framework which can allow deployment of fire models considering users’ requirements in terms of area and time. The framework allows integration of new fire models.

• A novel deadline based scheduling algorithm for efficient bushfire prediction.

• A case study using the Tasmania Bushfire Model for evaluating the Cloud based framework.

In the next section, we discuss requirements for a fire prediction service. Then in the subsequent sections, we describe the design and implementation of the proposed framework with evaluation and results. Then we discuss related work on fire prediction services and their comparison with the architecture of the proposed framework. Finally, we present conclusions and future directions.