Ukraine trains AI models by collecting combat data
Artificial intelligence, which is the most talked about technological development and has left its mark on the last period, has also become a part of the battlefields. In the Russia-Ukraine war, signs were revealed about how this technology was used and how it could be used in the future.
Used on the battlefields by both Russian and Ukrainian during Russia’s military intervention in Ukraine, AI became an important tool for identifying targets, especially with its ability to scan images much faster than humans.
DATA COLLECTION AND OCHI SYSTEM
Oleksandr Dmitriev, founder of a Ukrainian digital system called OCHI, also told Reuters on the subject, stating that they have collected 2 million hours (228 years) of war video collected by more than 15 thousand drone teams working in battlefields since 2022. Dmitriev described these images as “food for AI” and said, “If you want to train an AI, you give it 2 million hours of data, and it will create something from it.”
The OCHI system collects and analyzes video streams from Ukraine’s battlefields on a central platform and helps military commanders combine video footage of drone teams on a screen, presenting the activities on the battlefields as a whole.
It is emphasized that these images are stored for use in training AI systems for various military operations.
DAILY 5-6 TERABYTE VERY
Dmitriev stated that on average five to six terabytes of new data are added from battlefields daily, and that this data will greatly contribute to the training of AI systems in various military operations. This collection of images will be a valuable resource for training AI in a variety of military operations, such as target acquisition, combat tactics, and assessing the effectiveness of weapon systems.
Ukraine also collects drone and CCTV footage with Avengers, another system developed by the defense ministry. This system detects approximately 12,000 pieces of Russian equipment using AI recognition tools. By developing drone swarms, Ukraine is experimenting with executing commands from an AI system through an interconnected group of drones.
Drone footage collected from Ukraine’s battlefields has great potential to improve the decision-making ability of AI models on battlefields. This collection of data will be used to train AI in a variety of military operations, such as target identification, combat tactics, and assessing the effectiveness of weapon systems. Ukraine’s defense efforts are likely to continue to further develop AI technologies on the battlefield.