How AI Can Boost Disaster Response and Recovery
Published: 22 March 2021
Artificial intelligence enables rapid analysis of satellite images to assess damage and priority areas.
Building disaster-resistant cities is an important element in achieving the sustainable development goals adopted by the United Nations.
Data collected by the Office of US Foreign Disaster Assistance and the Centre for Research on the Epidemiology of Disasters for natural hazards across occurrences shows the number of disasters across the globe has risen by 74.5%, comparing data from 1980–1999 with 2000–2019.
The Intergovernmental Panel on Climate Change reports that the average global temperature has risen 0.85 degrees from 1880 to 2012 and is projected to rise further, affecting human society and the ecosystems.
Remote sensing technology using satellite images is now being used to investigate post-disaster situations, and to formulate reconstruction plans. In recent years, the number of commercially available satellites (e.g., Airbus, Maxar) has increased, and international charters have been created to obtain satellite images more easily during emergencies caused by natural hazards, such as earthquakes, heavy rainfalls, and wildfires.
The far-reaching impact of many natural hazards requires the rapid analysis of large numbers of images, which was conventionally done manually by people. This involves extracting from satellite images such information as damaged locations, sediment volume, number of victims, and building wreckage, which are crucial for post-disaster response.
Artificial intelligence (AI) is capable of rapidly analyzing large amounts of satellite images in a short period of time.
During the 2018 Hokkaido earthquake in Japan, AI was used for landslide detection. It took about 5 days for skillful engineers to distinguish between damaged places and misleading locations, such as farmlands and roads. Image interpretation by AI takes only 5 minutes to detect damaged places with an accuracy of 93% compared with human visual interpretation.
In the 2011 Tohoku earthquake, AI analysis was used to detect damages to houses. The work was carried out to understand the locations of houses that were washed away by the subsequent tsunami. The AI was able to spot the affected houses with an accuracy of 94%.
Remote sensing satellites are classified into two categories: optical satellites and synthetic aperture radar (SAR) satellites. An optical satellite provides almost the same images as human visual sense. Although visual interpretation of optical satellite images is relatively easy, the chance of acquiring these images is sometimes limited because of weather conditions, such as clouds and rainfall. On the other hand, a SAR satellite carries a radar sensor and acquires images in spite of weather conditions.
An example of SAR image analysis by AI is the detection of landslide caused by the 2020 heavy rainfall in Kumamoto. It is more difficult to distinguish the actual landslide from an optical image. AI analysis was able to detect landslide areas that were also identified by skilled engineers.
Immediately after a natural hazard, such as earthquakes, tsunamis, and heavy rainfall, it is crucial to assess the situation in a short period of time for rescue and countermeasures. AI analysis from optical and SAR satellite images is an important tool that can meet this need.
Central and municipal governments can benefit greatly from the use of AI and satellite images to alleviate damages from natural hazards and provide immediate relief to affected people.
Satellite images and AI analysis can assist governments in making quick decisions after natural hazards, including the identification of houses that have collapsed after earthquakes, assessment of damages to houses due to fires, and detection of flooded areas due to tsunamis.
A common issue in various types of natural hazards is the disposal of waste, such as dirt, sand, and collapsed infrastructure. Since the treatment of disaster waste is a heavy burden on the impacted area, the cooperation of central governments and the surrounding local governments is crucial for recovery. Estimating the amount of disaster waste by understanding the number and location of buildings that have been destroyed will aid governments and municipalities in reconstruction and waste disposal efforts. The ability of AI and satellite images to rapidly obtain information on objects and their locations allows for faster and efficient responses to natural hazards.
A. Fujita, et al. 2017. Damage Detection from Aerial Images via Convolutional Neural Networks. IAPR International Conference on Machine Vision Applications. Japan.
International Charter Space and Major Disasters. 2000. Charter on Cooperation to Achieve the Coordinated Use of Space Facilities in the Event of Natural or Technological Disasters Rev3.
Intergovernmental Panel on Climate Change. 2014. IPCC Climate Change 2014: Synthesis Report Fifth Assessment Report.
United Nations General Assembly. 2015. Transforming our World: The 2030 Agenda for Sustainable Development. New York: United Nations.
Leave your question or comment in the section below:
YOU MIGHT ALSO LIKE
The views expressed on this website are those of the authors and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.