Artificial intelligence-supported computer vision technology developed by one of the prominent companies in the defense sector in Türkiye directly helped search and rescue operations conducted during the earthquake disaster by determining damaged buildings, a deputy manager of the firm noted Sunday.
Enis Müçteba Memiş, deputy general manager of technology at Savunma Teknolojileri Mühendislik (STM), operating under the Presidency of Defense Industries (SSB) stated the determination process relating to the number of people trapped under debris in the earthquake-hit zone, and direction of search and rescue teams toward buildings that collapsed, was facilitated through an artificial intelligence (AI) program developed by the company.
Technologies developed by experts from STM were used effectively in search and rescue activities after the pair of deadly Feb. 6 earthquakes.
With AI computer vision algorithms, that work based on the comparison of satellite and UAV images taken prior to and after the earthquakes, developed by STM engineers, and with the support of images provided by the Ministry of Interior, and General Directorate of Population and Citizenship Affairs (NVI), it was possible to asses the damage to buildings in the earthquake-hit zone.
Using this data, rescue teams were directed based on information about residents of buildings in the Spatial Address Registration System (MAKS). Density maps were created for buildings' status and residents by matching the information obtained as a result of artificial intelligence studies via the MAKS-provided information.
In this way, accurate damage assessment was ensured and search-and-rescue teams were directed to wreckage sites.
Memiş, who provided information on the digital traces the company developed, noted that the Interior Ministry's MAKS system consisting of maps provides the identification number of how many people live on each street without information on buildings in areas.
"Through the live images of buildings we have established digital prints and integrated them into the MAKS system," Memiş added.
"Later, we conducted various analyses to guide teams to sites where buildings collapsed and to ascertain how many people might have been in each building with assistance from current images. At the moment, post-earthquake artificial intelligence software, which Türkiye can use during preparatory phases for new possible disasters, have been developed this way," Memiş underscored.
Pointing out that it is critical to develop these studies and works before disaster strikes, Memiş said, "There is information in MAKS about vulnerable cities. The geographical location of each building, and who lives in that building are known."
He also added that in this manner "digital copies" of cities can be made and estimation for possible earthquakes can be reached virtually as well.
"Accordingly, we can carry out search-and-rescue activities and engage in communication plans before the disaster. This system was originally developed as a base version of it. If this is combined with a large logistics information system, Türkiye will be better prepared for the potential next disaster," he concluded.