Evidence-driven disaster risk management (DRM) relies upon many different data types, information sources, and types of models to be effective. Tasks such as weather modelling, earthquake fault line rupture, or the development of dynamic urban exposure measures involve complex science and large amounts of data from a range of sources. Even experts can struggle to develop models that enable the understanding of the potential impacts of a hazard on the built environment and society.
In this context, this guidance note explores how new approaches in machine learning can provide new ways of looking into these complex relationships and provide more accurate, efficient, and useful answers.
The goal of this document is to provide a concise, demystifying reference that readers, from project managers to data scientists, can use to better understand how machine learning can be applied in disaster risk management projects.
There are many sources of information on this complex and evolving set of technologies. Therefore, this guidance note is aimed to be as focused as possible, providing basic information and DRM-specific case studies and directing readers to additional resources including online videos, infographics, courses, and articles for further reference.
A machine learning (ML) algorithm is a type of computer program that learns to perform specific tasks based on various data inputs or rules provided by its designer. Machine learning is a subset of artificial intelligence (AI), but the two terms are often used interchangeably.
For a thorough discussion of the differences and similarities of the terms ML and AI, see Section 2. As the name implies, an ML algorithm’s purpose is to “learn” from previous data and output a result that adds information and insight that was not previously known.
This approach enables actions to be taken on the information gathered from the data; sometimes in near real time, like suggested web search results, and sometimes with longer term human input, like many of the DRM case studies presented in this document.
Over the past few decades, there has been an enormous increase in computational capacity and speed and available sensor data, exponentially increasing the volume of available data for analysis. This has allowed the capabilities of ML algorithms to advance to nearly ubiquitous impact on many aspects of society.
Machine learning and artificial intelligence have become household terms, crossing from academia and specialized industry applications into everyday interactions with technology—from image, sound, and voice recognition features of our smartphones to seamlessly recommending items in online shopping, from mail sorting to ranking results of a search engine. The same technology is being leveraged to answer bigger questions in society, including questions about sustainable development, humanitarian assistance, and disaster risk management.
When several ML algorithms work together, for example, when fed by a large quantity of physical sensors, it is possible for a computer to interact with the physical world in such a way that the computer system, or robot, appears to be behaving intelligently. For example, self-driving cars, robotics that mimic and surpass human capacities, and supercomputers can now outperform humans on specialized tasks. The same expectation is, and should be, held for ML as it applies to improving our capacity to accurately, efficiently, and effectively answer pressing societal questions. The case studies in this guidance note range from the identification of hurricane and cyclone damage-prone buildings to mapping the informal settlements that house the most vulnerable urban populations.
For the understanding of disaster risk, machine learning applies predominantly to methods used in the classification or categorization of remotely sensed satellite, aerial, drone, and even street-level imagery, capitalizing on a large body of work on image recognition and classification. But applications also span other types of data: from seismic sensor data networks and building inspection records to social media posts. All the advancements made in the applications of ML can and are being used to solve bigger issues confronting humans, from making the most of our land to preparing for and recovering from crises.