Our university's Earthquake and Structural Health Application and Research Center (DE-YAS) is developing a fast, safe, and automated system using robotic dogs and artificial intelligence, which will replace traditional post-earthquake damage assessment methods. This project offers a pioneering approach in Türkiye and globally by enabling the rapid and automatic detection of damage in structures.

With this project, robotic dogs and artificial intelligence provide a safer, faster, and automated alternative to traditional observational damage assessment methods for buildings damaged in earthquakes. The process aims to accelerate interventions and repairs to buildings in the aftermath of disasters. Supported by KTU Coordination Unit of Scientific Research Projects and the Research Universities Support Program (BAP13), the project aims to be a first of its kind both in Türkiye and worldwide.

Our Project Will Support the Ministry of Environment, Urbanization, and Climate Change

In his evaluation of the project, our Rector, Prof. Dr. Hamdullah Çuvalcı, said: “Our country is located on a significant earthquake zone. Therefore, we must be prepared for earthquakes that could cause loss of life and property in every region of our country. Faculty members from KTU Department of Civil Engineering provided extensive support to the General Directorate of Construction Affairs at the Ministry of Environment, Urbanization, and Climate Change during the February 6, 2023, Kahramanmaraş earthquakes. We are launching this project to address the challenges encountered during this process and to support our ministry by accurately determining structural health.”

Risks in Damage Assessment Processes Will Be Reduced

Prof. Dr. Ahmet Can Altunışık, Director of KTU Earthquake and Structural Health Application and Research Center (DE-YAS), explained that the project aims to create a fast, reliable, and automated decision support system as an alternative to the traditional observational damage assessment methods used in buildings affected by earthquakes. Prof. Dr. Altunışık summarized the project's goal as follows: "With the help of robotic dogs used in the project, high-resolution images will be collected from the surroundings and interiors of damaged buildings, human intervention will be minimized, and these images will be analyzed using artificial intelligence algorithms to contribute to damage assessment efforts. Especially after the February 6, 2023 earthquakes, damage assessments were carried out in approximately 2.6 million buildings, with 10,000 technical personnel involved, and 5-10% of the buildings were re-examined. In this context, the importance of the decision support system developed within the project becomes even more significant. Additionally, thanks to the robotic dog, it will be possible to avoid the need to enter buildings and work under risk during aftershocks.”

A First in Türkiye and Worldwide

Project Executive Prof. Dr. Kemal Hacıefendioğlu stated that the project is a global first in terms of both the methods used and the solutions it offers. Prof. Dr. Hacıefendioğlu made the following remarks about the project: “Traditional damage assessment methods can be time-consuming and risky as they rely primarily on human labor. However, within this project, it is aimed to inspect the exterior and interior of structures quickly and safely with the help of robotic dogs. By analyzing the images with artificial intelligence algorithms, the type and severity of structural damage will be automatically determined without the need for human intervention. Thus, human error will be minimized, the damage assessment process will be accelerated, and more accurate results will be obtained for our ministry.” The software to be developed within the project will offer a significant innovation by fully automating damage assessment reports. Since robotic dogs can collect data in challenging terrains and inside dangerous structures, this technology will enable a more in-depth analysis of the damage and make the damage assessment process much more efficient and safer.”

Deep Learning Algorithms Will Determine the Extent of Damage

KTU DE-YAS Deputy Directors Assoc. Prof. Dr. Hasan Basri Başağa and Asst. Prof. Dr. Fatih Yesevi Okur emphasized that the innovative aspect of the project is based on analyzing the images collected by robotic dogs using deep learning techniques. In the first phase, the robots will roam around the damaged structures and collect external images. If the external data are insufficient, the robotic dogs will enter the interior spaces of the building to gather additional images. These images will be analyzed by artificial intelligence algorithms to determine the type and severity of the damage. The deep learning algorithms used in the project will be trained on large data sets, enabling the detection of the damage with great precision. In the second phase, the software developed within the project is planned to be integrated into the Ministry of Environment, Urbanization, and Climate Change’s existing damage assessment system. This will accelerate post-disaster building assessments and speed up the repair processes of the structures.

The project team extended their gratitude to the Rector of our University, Prof. Dr. Hamdullah Çuvalcı, and the President of the Council of Higher Education (CoHE), Prof. Dr. Erol Özvar, as well as the entire KTU Scientific Research Projects team for their support. They also expressed their hope that the project would contribute to the development of science and technology in Türkiye.

 

Önceki Sonraki