A Look Around Innovation, Part 33: Rapid Earthquake Damage Assessment System Using Sensors — Jointly developed with Peru, utilization begins in disaster prevention drills

In the segment 'A Look Around Innovation,' we introduce research and development (R&D) sites that have led to real-world applications. In the 33rd installment of this series, JST News interviewed Professor Koichi Kusunoki of the Earthquake Research Institute at The University of Tokyo, who led an international joint research project with Peru, a country frequently affected by earthquakes. The project developed a system that uses sensors to instantly assess earthquake and tsunami damage, helping support emergency response efforts immediately after a disaster. By rapid estimation and observation of Damage Level of buildings, roads, and water infrastructure the Integrated Expert System is expected to support emergency response efforts immediately after an earthquake and facilitate faster recovery. The system is already being used in disaster preparedness drills in Peru, and the findings are also expected to contribute to earthquake and tsunami countermeasures in Japan.

Integrating information through digital technologies to support administrative initial responses

Nestled between athletic fields and a baseball diamond on The University of Tokyo's Yayoi Campus in Bunkyo City, Tokyo, stands the Earthquake Research Institute. Established in 1925 in the wake of the Great Kanto Earthquake of September 1, 1923, the institute has a history spanning more than a century. At the institute, Kusunoki is conducting international collaborative research with Peruvian universities, research institutes, and government agencies.

Like Japan, Peru, on the opposite side of the Pacific, continues to suffer from earthquakes and tsunamis. Through initiatives such as the Japan-Peru Center for Earthquake Engineering Research and Disaster Mitigation (CISMID), established at Peru's National University of Engineering in 1986, Japan and Peru have long collaborated on research and outreach activities related to urban disaster management planning and disaster mitigation technologies. Together with Professor Carlos Zavala of CISMID, Kusunoki developed SEIDAS, a system that automatically assesses disaster damage in Lima, Peru, using sensor technologies through JST's SATREPS program. SEIDAS not only detects seismic shaking and predicts tsunami impacts but also assesses damage to buildings, sewer systems, roads, and other infrastructure, helping authorities determine appropriate response measures. "What needs to be done immediately after a disaster is to quickly determine the scale of the event and identify where damage has occurred, then launch an initial response," Kusunoki explains.

In the immediate aftermath of a disaster, confusion is widespread and information about damage is often scattered across multiple sources. One of the project's key goals is to overcome this challenge through digitalization and information integration. "The disaster management field was slow to embrace digitalization. For a long time, there was a belief that reliable analog technologies should be prioritized because power and communication networks often become unavailable during disasters. Today, however, electricity and internet services can be restored much more quickly. That led us to conclude that the time had come to develop disaster response systems built around digital technologies," he recalled, explaining what led him to begin the project.

SEIDAS can be used not only to assess actual disaster impacts but also as a platform for simulation-based disaster preparedness exercises. When users enter an earthquake's magnitude, the system displays projected tsunami inundation zones, building-collapse risks, and potential damage to roads and water infrastructure on a map (Figure 1). Area-specific reports also provide estimates of how many days will be required to restore water service. These outputs can be incorporated into routine disaster preparedness drills.

Calculating collapse risk for faster emergency response: AI-powered damage assessment in development

The international collaborative research project with Peru consists of four research themes (Figure 2). Among them, Kusunoki led the research group focused on buildings under Research Theme 2.

In Japan and elsewhere, assessing the condition of damaged buildings typically requires inspectors to visit sites and conduct visual surveys. Kusunoki's approach, by contrast, is to install accelerometers (Figure 3) in buildings and connect them to a network, enabling the building's shaking, deformation, and risk of collapse to be assessed immediately after an earthquake. The key is that the system does not perform all calculations from scratch when an earthquake occurs. Instead, it compares shaking data detected by sensors with a vast database of damage scenarios that have been precomputed and accumulated in advance. This makes it possible to estimate collapse risk almost instantaneously and provide valuable information for rapid initial response efforts.

Kusunoki emphasizes the simplicity of the system: "No more than four or five sensors are typically needed to detect a building's motion. People often ask whether building blueprints are required or whether a computer model of the building must be created in advance. Neither is necessary. As long as all sensors are installed horizontally and time-synchronized throughout the building, earthquake-induced shaking and deformation can be detected. To assess collapse risk, we also take into account structural characteristics such as construction type, whether a building is wooden or brick-built, its number of stories, and its age." To synchronize the sensors' clocks to within one-hundredth of a second, a small computer is installed inside the building.

The team is also developing artificial intelligence (AI) technology that can estimate damage levels in buildings without installed sensors by using measurements taken from nearby buildings together with information on structural characteristics. At the same time, Kusunoki cautions that the role of AI in disaster response must be considered carefully. "Errors and uncertainties are inherent in AI. When human lives are at stake, we need to think very carefully about whether such errors and uncertainties are acceptable."

Mitigating damage from heavy rains, typhoons, and other disasters — Applying the findings to disaster research in Japan

At the end of May 2026, Peru conducted a nationwide disaster preparedness drill, during which SEIDAS was demonstrated in Lima's Chorrillos district. Prior to the exercise, Kusunoki visited the site and explained the functions and operation of SEIDAS to disaster management personnel. Looking ahead, he hopes to expand the system's coverage to all of Peru and eventually to other parts of Latin America. To achieve this, he says the first step will be gathering basic information on buildings, roads, and other infrastructure. Kusunoki also hopes to promote the digitalization of disaster response, including the use of SEIDAS, at the 2027 Asia-Pacific Ministerial Conference on Disaster Risk Reduction, hosted by the United Nations Office for Disaster Risk Reduction, scheduled to be held in Sendai.

Currently, Kusunoki also serves as Program Director for the Cabinet Office's Strategic Innovation Promotion Program (SIP) initiative, "Development of a Resilient Smart Network System against Natural Disasters." By scaling up the achievements of the SATREPS project, he aims not only to improve responses to earthquakes and tsunamis but also to help reduce damage from heavy rain, typhoons, and other natural hazards while enabling faster recovery. "The timing of Japan's push toward the digitalization of disaster response coincided with our efforts to advance digitalization together with Peru, making SATREPS a highly timely international collaboration. The significance of this project lies not in transferring technologies developed in Japan to Peru, but in creating something new together with Peru that did not previously exist in Japan," he says. Kusunoki added that personnel exchanges and the transfer of knowledge to younger generations are also progressing and expressed his hope that these outcomes will contribute to future disaster preparedness research in Japan.

(Article: Shosuke Shimada, Photography: Erika Shimamoto)