The Flight Safety Service (FSS) of the Russian Ministry of Defense together with Rostec Subsidiary, JSC RT-Techpriemka plan to introduce advanced machine learning in the production and operation of military aircraft.
The use of artificial intelligence will make it possible to carry out long-term planning of product quality and predict the need for equipment repair with high accuracy. A new approach to collecting and analyzing data will increase forecast accuracy by up to 90%, Rostec said in a release.
“We are positive about the idea of using modern technologies, including artificial intelligence. The availability of objective and complete information about the root causes of possible failures of aviation equipment at the level of technological processes will help prevent more serious events. After all, where the technological process is built, safety is ensured,” said Sergei Bainetov, Head of the Aviation Safety Service of the RF Armed Forces.
The implementation of advanced machine learning is designed to minimize safety risks. The method is based on the use of predictive analytics - with its help it is possible to predict the behavior of technology in the future. For example, even at the assembly stage, the system is able to identify a potential defect that can provoke equipment failure. During maintenance or repair, the solution will detect factors affecting the state of the equipment and predict possible emergency situations, allowing them to eliminate their causes in advance.
To create an innovative forecasting system, large-scale work will be carried out to collect and digitize production indicators. In particular, the volume of production, the parameters of technological processes, the speed and quality of processing of complaint events will be analyzed. Based on the received data, the "smart" system will study the production and operation processes, calculate their optimal parameters and will be able to issue recommendations to the operator. Machine learning technology will allow you to create a mathematical model of the production process, which can then be integrated into an automated control system.
“The introduction of artificial intelligence will make it possible to solve many issues at all stages of the life cycle of military aircraft, starting from the moment of design. Thanks to new technologies, maintenance work is optimized, downtime will be reduced, and the operator will have complete information about the condition of the equipment and, most importantly, will be able to prevent equipment failures. Now we are on the verge of a large-scale study, on the basis of which the model will be trained. To achieve maximum accuracy, it needs to be trained in several stages, adding or removing some datasets. The final product should be a web service or a mobile application with a user-friendly interface,” said Vladlen Shorin, Director General of RT-Techpriemka JSC.
On April 9, U.S. officials said AI must receive at least 3.4% of the DOD budget. Marine Corps Lt. Gen. Michael S. Groen, director of the Joint Artificial Intelligence Center, told reporters at the Pentagon, “AI isn't just important for the DOD's warfighting capabilities, it will also be a powerful driver of the American economy.”
Another top official, Robert O. Work, vice chair of the National Security Commission on Artificial Intelligence (NSCAI), said that identifying the right talent for AI, ensuring AI is used in an ethically responsible manner and forming international partnerships are other ingredients for successful AI implementation.