Samenvatting

''Ever increasing population in seismically active urban areas, aging building stock, and expansion of urbanization to previously agricultural lands with soft soil deposits render the protection of human lives against earthquake disasters extremely more difficult by the time. Although much effort is put in further improving the current seismic design practices for new buildings, recent earthquakes show us, again and again, that life losses occur in older and much more vulnerable structures. Finding those substandard, collapse-vulnerable buildings before a destructive earthquake is like finding a needle in a haystack. It is clear that the problem in hand cannot be addressed with the existing, and mostly old-fashioned tools anymore. This manuscript focuses on how the emerging technologies, such as Artificial Intelligence, image processing, and data sciences in general, can be implemented as useful tools for conducting an urban scale seismic risk assessment while estimating the risk for every individual building. A review of the available technologies is given for the exposure component. Furthermore, a novel method of estimating the vulnerability of individual buildings, based on autoregressive machine learning algorithms, is presented. The manuscript discusses that the technological advancement is mature enough to radically alter how the earthquake risk is estimated.''