Information Driven Evacuation System (I.D.E.S.)
Grindrod, Samuel Edmund
The effectiveness of an emergency response during an incident is often affected by the lack of information provided to the people within the situation about the current conditions. Deaths in large-scale fires are often likely to have been caused by delays in the occupants receiving relevant information on the fire and egress routes. This is why pre-movement behaviour, which is defined as the behaviour which occurs before an alarm is sounded and includes the activities which occur between the alarm sounding and the occupants beginning to move towards an exit, is believed to be generally more important to survival than the actual movement speed. It is the unpredictability and complexity of human behaviour that is the most influential factor on the success / failure of an evacuation plan. Unfortunately, evacuation plans rely on the use of purposely designed egress routes which often are not the common everyday exits. These specifically designed egress routes, which an engineer may assume will be used during an evacuation, are often ignored by occupants due to the lack of information and noticeable distinguishing features. Having occupants moving in directions away from these intended routes may result in the increasing possibility of occupants finding themselves in a dangerous situation, ultimately leading to potential loss of life. The value of a sensor-linked fire model has been demonstrated and the potential for interpretation of human behaviour shown. However, there are many challenges in representing and interpreting data on human behaviour. Within most emergency evacuation situations, occupants will often walk past emergency exits without using them and exit through the main entrance or main exit, as displayed during an evacuation experiment held in IKEA in 1996. Problems occur because occupants will rely on the familiar exits over the closest emergency exit, which could be potentially overcome by the use of an information driven evacuation system. The main function of the Information Driven Evacuation System or I.D.E.S. is to provide occupants with information on the most appropriate egress paths within a building based on the development of the fire and the movement of other occupants. The system is a combination of real-time sensor data, a prediction modelling tool and the information driven way-finding tools. However, as all three processes are independent systems, a central server will be required in order to ensure that all the different processes are speaking the same language and that the information from one system can be understood by another. of the components within the system interact with each other. The basis of the system will combine the use of sensors within a building and specific way-finding tools to give the I.D.E.S. the ability to change the information provided by the way-finding tools by having the sensors within the building interfacing with a computer server. This server will incorporate a modelling program that will have ability to assess the data gathered by the sensors, and use the servers “intelligence” (i.e. predicting capabilities) to alter the information provided by the way-finding tools. The server will also have the ability to use the sensor data to predict the development of the fire and the movement / behaviours of the occupants. The way-finding tools used within the I.D.E.S. would have the primary goal of relaying the information to the occupants within the building through the use of both audio (e.g. directional speakers) and visual (e.g. flashing lights) capabilities. Basic audio and visual tools are already used as common features of an evacuation plan  and include exit signage and alarm bell/sirens. The computer model used as part of the “intelligence” of the server will need to have predicative capabilities that incorporate information provided in real time. It is believed that the combination of these tools will be able to provide the occupants with the information required to evacuate the building in a safe and efficient way without causing confusion, thus reducing the possibility of stress and anxiety. However, the solution will only work if the combination of the tools, sensors and systems are able to be integrated into a central control panel that can be understood and used effectively by fire service and/or security staff. The following is the Chapter breakdown of the thesis: Chapter 1 discusses the nature of the problem that is to be addressed by the I.D.E.S. as well as the proposed solution and the overall concept of the system. Chapter 2 provides an overview of the system to be developed as part of this thesis, via a graphical overview as well discussing the current status of the system and an higher level summary. Chapters 3, 4 and 5 discusses the background research and information gathered on the current code requirements for an evacuation design, the current theories and completed research focusing on the human behaviour of occupants during an evacuation, and the development of computer modelling programs and their historical implementation. Chapter 6 focuses and describes the development of predictive modelling, the limitations that dictates the modelling process, the requirements of the I.D.E.S. modelling programme and a review of the possible programmes for their ability to be used as part of the system. Chapter 7 covers each of the three evacuation experimental series that were conducted as part of this thesis. The purpose of each of the relevant experiment is covered, the results gathered, the analysis of the research, and how the results influenced the development of the simulation methodology. Chapter 8 focuses on the development of the modelling programme, based on the experimental data gathered, and presents a feasibility study that will demonstrate how the system would work during a simulated real life evacuation based on the information gathered from the experiments using the updated CRISP program and the hypothetical installation of the system within an existing building, located in Auckland, New Zealand. The possible further evolution of the system is discussed in Chapter 10.