Handbook of Cognitive and Autonomous Systems for Fire Resilient Infrastructures

The framework for performance-based design for fire safety that is in use today is based largely on constructs that emerged in the early 1990s. The framework has its origins in systems approaches to fire design for buildings that were pioneered in the 1970s, which in turn made use of the fire safety science principles and constructs that began to emerge in the 1950s. It has proven to be adaptable to deterministic and probabilistic realizations, and is arguably a risk-informed approach, whether benchmarked to tolerable risk as embodied in regulatory provisions or makes use of quantitative risk measures. The framework contemplates technologies—in the form of safety technologies and computational modeling for hazard assessment—and people—primarily as targets to be protected by the safety technologies. The framework also considers the regulatory environments within which it is applied. Nonetheless, performance-based design for fire safety is not as broadly accepted as performance-based design approaches in other disciplines. Arguably, this is due in part to a lack of a socio-technical systems framing and due consideration of the associated people-technology-institutions interactions that impact fire safety throughout the life of a building. Stakeholders have concerns about the application of technologies in the design process, the qualifications of practitioners, and how the building will perform in the future. Furthermore, current approaches to design often do not incorporate the technologies that can help maintain a target level of fire safety performance, either by notifying persons who can take action, or autonomously modifying building fire safety parameters. These challenges can be overcome. This chapter introduces some concepts of socio-technical systems thinking and system safety thinking, how they can be applied throughout the lifecycle of a building, and how these concepts and approaches can result in more robust, sociotechnical systems oriented, performance-based designs for fire.

Most construction types are adversely affected by fire, if it occurs. If the construction is combustible, then the structure may burn down. But even if it is not combustible, a serious fire, especially a post-flashover fire, may result in major damage or even collapse. This is because few materials are available which can stand prolonged application of high temperatures without degradation or failure.

Today we are stepping into the new age of Cyber Physical Systems (CPS). Dramatic technological advances are providing novel opportunities for the safety infrastructure to achieve high-reaching levels of resilience to adverse events.

Fire resiliency is dependent on built-in fire protection measures and manual intervention techniques. Built-in fire protection measures utilize passive and active systems and similar approaches. Manual intervention techniques are implemented through the emergency responders such as the fire service, occupants, and others, who use equipment and operations to control and mitigate unwanted emergencies.

Our world is rapidly evolving to enable Cyber Physical systems that will become fused with other building systems. These will effectively collect, process, and deliver (i.e., act-on) data, and will revolutionize fire protection and emergency response. The result will have significant potential to support efficient societal improvements and a fire-resistant infrastructure for our future world.

Building’s fire prevention strategy tailored with the creation of urban buildings design has developed rapidly while promoting the more advanced modern cities. After nearly a hundred years of experiences in fire protection, the concept of building’s fire protection of the so-called active fire protection design and passive fire protection design has gradually been reformed. With the development of high-speed computer technology and the improvement of modern science and technology, a new concept of the so-called performance-based fire protection design could be first traced down as early as 1970s when the goal-oriented approach to building fire safety was developed by the U.S. General Services Administration. The design concept uses the computational fluid dynamic (CFD) software to obtain the movement data and the trace of fire and smoke, temperature distribution, gas concentration’s distribution cloud map, and other parameters. The emergence of this new technology enlarged the traditional prescriptive-based fire protection design approach with the alternative design method if code complaint approach does not fit the design best. The fire protection design of complex buildings can use this technology to perform the optimized design or justification / verification analysis quantitatively and obtaining more sound results and conclusions. However, the focus is still on how to fight and control the fire while preventing the fire as the second. With the new performance based design (PBD) method, the fire risk could be mapped and quantified in order to have a better weight in analysis toward a better fire and life safety tailored design. This could be even more achieved with the new tools including the information technology, Internet of Things, cloud computing, Big Data, and artificial intelligence. The PBD approach has been used in other engineering disciplines for longer time and approved to be a scientific based sound approach, with the new empowered information cutting-edge technology captioned above, fire protection design can take the advantages without leaving behind. "Smart Building" is a new thing that is being nurtured in the current era. It is a building that provides safety, efficiency, comfort, and consumes less energy for building, people, and environment by using computer technology and Internet technology. Under the ideal condition, smart buildings are supposed to be safe and the level of fire safety can meet people's expectations based on the reliable risk management tools. The implementation includes how to recognize the building features to identify the risk levels and areas, sending signals alerting the building’s brain for further diagnosis and action. The smart building concept is getting more and more attention with penetration and coordination of various engineering disciplines. The topics we are discussing in the chapter is aimed to be a food for thoughts making a better fire protection design and ultimately a better safer world.

The concept of Smart and Autonomous Infrastructure (SAI) and the legacy of building/fire codes remain in philosophical conflict. Specifically, prescriptive building/fire code provisions are structured in a layered or defense-in-depth fashion rather than a holistic and synergistic manner. Consequently, many designers have been conditioned to evaluate systems in isolation with little thought of their synergistic potential. Also, these provisions are primarily based upon past experience or precedent and collectively represent an immense barrier to entry by new technologies. However, developments in performance-based design of buildings and infrastructure constitute viable outlets for holistic and emerging technologies such as SAI. Existing performance-based design methodologies provide owners and stakeholders with quantitative information to make risk-informed decisions for facilities exposed to multiple hazards (e.g., fire following earthquake) and design features based on predefined performance goals (e.g., life safety, continued functionality). Conceptually, such methodologies include components that can be readily adapted to permit the incorporation of SAI concepts. Ultimately, the implementation of SAI into building design could result in designs with enhanced reliability to (1) safeguard the public health, safety, and general welfare of buildings occupants, fire fighters, and emergency responders, as well as (2) mitigate damage to property.

Over the past decade, big data and artificial intelligence (AI) enable new smart techniques in the building and construction area. The applications of AI in fire detection, risk assessment, and fire forecast are emerging. This chapter provides a roadmap for AI-based building fire safety engineering application by comparing it with the history of CFD fire modelling. Guidelines for constructing a reliable fire database with both experimental and numerical data are introduced. The AI algorithms having a great potential to detect and forecast fire scenarios are discussed, and the latest research on exploring and developing intelligent firefighting systems are reviewed. Finally, three new concepts of applying AI in building fire safety are proposed, (1) the AI-based fire engineering design to improve the structure fire safety, (2) the building fire Digital Twin to monitoring the fire risk and development in real time, and (3) the Super Real-time Forecast (SuRF) of the fire evolution.

Firefighting is a dangerous activity that puts firefighters into conditions that threaten their safety in order to save lives. To reduce firefighter injuries and limit their exposure to hazardous conditions, a range of technologies have been developed to improve their planning, situational awareness, and firefighting activities. These technologies allow firefighters to be more intelligent about their activities to reduce the likelihood of injury and be more effective. This chapter provides an overview of technologies currently being used as well as those being developed to support more intelligent firefighting. This includes monitoring devices, imaging systems, and robotic platforms.

Artificial intelligence (AI) as a discipline focuses on developing computer systems which can interpret raw data to make observations, interpretations, and decisions which traditionally have required human insight. There are numerous decisions which need to be made on the fire ground during structural fire suppression operations. What specific hazards exist, where are the exposures, what is the most effective method of extinguishment, and does the survivability profile warrant interior rescue operations to name a few. Due to the restrictive timeline of a fire, these decisions are often based on limited information. As buildings are designed with more intelligent systems, it will be even harder for emergency responders to analyze all the available data in making these decisions without relying on AI-based systems. This chapter provides an overview of AI, available data, and the benefits AI systems can bring in data interpretation and decision-making in the fireground.

Achieving situation awareness (SA) among first responders is an enduring challenge. As illustrated perhaps most starkly in the 9/11 response to the World Trade Center, the consequences of separate command posts and inability to utilize radio interoperability led to a breakdown of SA which exacerbated the toll in terms of deaths, primarily among first responders ([1]; [2], p. 9).

The fire-resistance testing has its long history. The early attempt of fire testing was based on exposure of the buildings assembles on fire. It was generally large-scale fire conditions in which full-size buildings elements were burned and evaluated. One of the first furnace tests is credited to T. Hyatt. He conducted tests in a wood-fired furnace, over which a specimen was tested. The test specimen represented sections of a floor slab. The section was loaded with a mass 1500 kg/m² for testing its bearing behavior. Then after 10 h, a hose stream test was conducted. Hyatt was unable to measure and control the temperature of furnace. However, with using the immersion calorimetry the results were surprisingly repeatable in obtaining the iron temperature.

This chapter focuses on advanced solutions for smoke control in buildings—active pressure differential systems (PDSs) used to maintain smoke-free evacuation routes. The discussed system relies on live measurements (pressure, temperature, and flow) and status updates from other building automation systems. Active systems provide improved operation compared to traditional solutions, require less resources, and if designed correctly less prone to cause dangerous interactions in their operation. The downside is the complexity of the solutions, which requires rigorous certification, factory acceptance testing, and commissioning. Furthermore, these systems require significant autonomy from other building automation systems, as data transfer and incompatibility with other building automation systems may lead to unexpected failures. An overview of modern solutions with varying complexity and capabilities is given, with some research gaps identified for future research.

The world’s present and future challenges are impacted by the climate change, more and more limited access to resources along with the increase and aging of population. The engineering world will adapt to these challenges by developing new materials, new type of structural solutions, or new fire-based evaluation tools. The new developed materials or structural elements must be tested in order to validate and calibrate the existing or new fire-based evaluation tools, and hybrid fire testing is a new emerging promising technique. Generally, the fire tests are performed on single members, under constant mechanical boundary conditions, neglecting the action of the remainder structure. However, full-scale testing of full structural systems revealed different behaviors of the structural members, but the high cost makes the practice uncommon. Hybrid fire testing is keeping the advantage of testing only parts of the structure, but at the same time considering the effect of the remainder structure, thus reproducing the results of the full-scale testing. In the context of new emerging fire-based evaluation tools, i.e., machine learning has the potential to facilitate performance-based fire design of structures. This chapter conceptually describes the overlap of the hybrid fire testing methodology and the new fire-based evaluation tool, machine learning. A detailed description of hybrid fire testing method is presented along with a benchmark case study for exemplification purposes.

For decades, the community of fire safety engineers have been dedicated to applying appropriate measures to improve the fire resistance of modern structures. However, estimation of structural resilience in fires is almost narrowed down to the practice of using standard fire curve and prescriptive design (i.e., applying fire protection simply according to the rate of fire resistance). Even with the recent spread of performance-based design (PBD) approaches within the fire safety community, the tools for and understanding of structural fire engineering still lag severely behind. In the 1990s, the ground-breaking Cardington fire tests have demonstrated the significance of investigating system-level responses in evaluating the fire resistance, from which the floor system was found to survive in a post-flashover fire through a tensile membrane action. A similar breakthrough was made after rigorous investigations on the tragic collapse of World Trade Center (WTC) buildings on September 11, 2001, which, by contrast, has shown the system-level vulnerability of modern designed structures. Had the towers not been hit by the aircraft and only set on fire, comprehensive investigations showed that they would have still collapsed. With the decades of research, many fundamental mechanisms of structural behaviour in fires have been identified. The advance of modern technologies and techniques has been utilised in understanding the structural behaviour in fires and ultimately preventing the fire induced failure and collapse. This chapter begins with a brief introduction of these fire induced collapses to underline the complexity of fire–structure interaction analyses, which is followed by a summary of the latest established design fire scenarios and structural failure mechanisms. While highlighting visionary application of autonomous infrastructure in evaluating structural fire resistance, smart technologies such as artificial intelligence (AI) have been summarised in the context of predicting fire behaviour and structural responses. As a pioneering attempt to investigate the collapse criteria for early warning, the tests in Sichuan Fire Research Institute (SCFRI) are presented to demonstrate the use of advanced technologies to monitor the critical events leading to fire induced structural collapse. At the end of this chapter, the prospect of future structural fire resistance evaluation is discussed, while such a vision will always focus on improving the safety performance of built environment considering fire threats.