Assessment of earthquake preparedness of existing structures: the case of Erzincan province

Evaluating and analyzing the existing building stock in cities like Erzincan, where the earthquake risk is remarkably high, is critical for urban planning and disaster risk reduction efforts. Although street scanning, one of the traditional field-based evaluation methods, has been used for a long time to collect data on structural features, these methods have significant disadvantages such as being time-consuming, labor-intensive, and open to evaluator-dependent errors. In this study, to evaluate the earthquake risk of the structures in Erzincan city center in a fast, objective, and effective way; two different traditional street scanning methods and an artificial intelligence-supported analysis module were combined in a single software platform. The survey approach of the ITU Alumni Association Bursa Branch was used as the first method, and Erdem Erdoğan's visual street scanning method was used as the second method. In the third method developed in addition to these, an artificial intelligence-based visual analysis module that can analyze photographs taken from four different facades of the structure using various models was utilized. The software provides the opportunity for fast and easy comparison by offering suggestions after performing analysis, evaluation, and scoring in separate tabs by combining three different methods in the same interface. As a result of the tests conducted in the center of Erzincan province, it was determined that the risk scores and categories produced by the artificial intelligence model showed 80-90% similarity with traditional methods. While the error margin of the software varies between 5% and 15%, the enormous time advantage it provides in processing speed is noteworthy. The analysis process of ten buildings, which takes hours with traditional methods, was reduced to approximately ten minutes with the software. This study demonstrates the potential of artificial intelligence in civil engineering applications and offers a fast, consistent, and reliable methodology for disaster management and building inventory updates