Development of teaching programs of artiﬁcial intelligence methods in aerospace education

. As a part of the National Strategy on Artiﬁcial Intelligence development in the Russian Federation, the basic concepts and deﬁnitions related to arti-ﬁcial intelligence (AI) and the necessity for the preparation of training programs were analyzed. The analysis of Federal state educational standards (FSES) and professional standards for enlarged groups of specialties and directions (EGSD) 24.00.00 AEROSPACE ENGINEERING was conducted and the perspective programs on the foundations of artiﬁcial intelligence and intelligent data analysis were proposed.


Introduction
The relevance of the considered task is determined by the Presidential Decree No. 490 of October 10, 2019 "On the Development of Artificial Intelligence in the Russian Federation".
The educational technologies should also rely on the Federal Law of 29.12.2012No. 273-FZ "On Education in the Russian Federation", the Federal standards of higher education concerning the professional standards of the consolidated group of training areas and majors EGSD (enlarged groups of specialties and directions) 24.00.00Aviation and Rocket-Space Engineering.
The application of artificial intelligence (AI) methods in education include two main trends: • use of AI methods in the organization of the educational process; • teaching of AI methods.The first trends is universal.Technologies such as adaptive learning, personalized learning, automatic assessment, interval learning, evaluation of the faculty by students, smart campuses, control of the examination process are currently under development.Appropriate information platforms are also being developed for organizing the educational process using new technologies.
An enlarged classification of the AI technology application includes the following aspects: • computer vision technologies in education (for monitoring students during online proctoring); • technologies of natural language processing, speech recognition, and synthesis; • technologies of intellectual support for decision-making.
The second trends of implementing AI methods in the curriculum relies on both the latest achievements in AI and the specifics of educational activities.
In articles [1,2] general approaches to the formation of educational programs are considered.The problems of digital transformation at a new stage are considered in [3,4].In [5][6][7][8][9][10], descriptions of methods of using intelligent technologies are given.Foreign experience is analyzed in the article [11].

Key concepts and definitions in the AI field used in the curriculum development
The key term, "artificial intelligence", is defined as "a set of technological solutions that allow the simulation of human cognitive functions, obtaining results comparable, at least, to the results of human intellectual activity".An expanded definition of AI is as follows: AI is the science and technology of creating intelligent machines, especially intelligent computer programs.AI is related to the similar task of using computers to understand human intelligence but is not necessarily limited to biologically plausible methods.The goal of AI is to create technical systems capable of solving non-computational tasks and performing actions that require the processing of meaningful information and are considered the prerogative of the human brain.
The main attributes and distinctions of intelligent systems (IS): 1. Developed communicative abilities-processing arbitrary requests in a dialogue in a language closest to natural language (the system of natural language interface-SNLII).
2. Focus on solving weakly structured, poorly formalized problems (implementation of soft models).
3. Ability to work with undefined and dynamic data.
4. Ability to develop a framework and extract knowledge from the accumulated experience of specific situations.
5. The ability to obtain and use information that is not explicitly stored, but is derived from the database.
6.The system has not only a model of the subject area but also a model of itself, which allows it to define the scope of its competence.
7. Ability to draw conclusions by analogy.
8. Ability to explain its actions, user failures, to warn the user about some situations that lead to the violation of data integrity.
In contrast to conventional analytical and statistical models, IS allows obtaining a solution to poorly structured problems that are difficult to formalize.
The main types of intelligent systems are: 1. Systems with intelligent feedback and intelligent interfaces.
7. Genetic algorithms and evolutionary modeling.
8. Knowledge extraction from experience (empirical facts) and Data Mining.
According to the above list, data mining is included in the list of intelligent systems.An alternative concept is the interpretation of Data Mining, which includes AI methods and is defined as a decision support process based on the search for hidden patterns in data.
The purpose of pattern finding is to present data as a representation of the searched processes.Building prediction models is also the goal of pattern search.Data Mining tools, unlike statistical tools, do not require a strictly defined amount of retrospective data.At the same time, it is necessary to control the statistical significance of the detected knowledge.In addition, 60 professional standards in the field of Rocket and Aerospace Industry were analyzed.AI General Course was recommended to be included in the training system for 22 of considered professional standards.
Analysis of 16 professional standards in the Aviation field showed the advisability of studying the general course of AI methods for at least four standards.

Development of the basic program on the Artificial Intelligence methods, recommended in the EGSD 24.00.00 educational programs
The main document to be developed is the educational program: a set of basic properties of education (volume, content, expected outcomes) and organizational and methodological conditions.This set is presented as a curriculum, a calendar of the curriculum, working programs of subjects, courses, disciplines (modules), other components, assessment and methodological materials, and in cases prescribed by Federal Law No. 273-FZ of 29.12.2012"On Education in the Russian Federation", as a working program.
The following program is offered to form a general basic course on AI methods.The course consists of modules that may be considered both as a whole and as separate components.
The specifics of the course are fulfilled through the consideration of end-to-end examples in lectures and practical classes.
The objectives of the discipline include the formation of a system baseline representation and primary knowledge, skills, and competencies in the AI basics and methods, along with insights into the implementation of artificial intelligence methods in technical applications.This course prepares teachers and students to apply the concepts of intelligent systems in scientific and practical areas.
The main tasks of the discipline are to acquire skills in the research and development of artificial intelligence systems in engineering practice.
The specific problems of implementing artificial intelligence technologies depend on the focus of a particular educational program.Therefore, a significant element here is an educational module-a set of educational elements sufficient to organize and pass the training on a particular topic (class) of the discipline (course).The educational module outcome is the development of certain knowledge, abilities, skills, or competencies by the student.
Description of the modules of the general course on AI methods is presented in table 1.According to the analysis of FSES 3++ for EGSD 24.00.00 and relevant professional standards in the fields of Rocket and Aerospace Industry and Aviation, we recommend including the "Intelligent Analysis of Data-DATA Mining" course for all training programs.
DATA Mining skills will become the basis in solving the most important problems for all engineers designing new equipment, specialists involved in the equipment operation, research engineers, and scientists.The received knowledge and skills will be irreplaceable for design decision-making support, prognosis, choosing the development directions, Big Data processing, and problem-solving in the limited information conditions.
An analysis of the problems in the aerospace industry resulted in the following course on data mining, presented in table 2.

Conclusions
1.The analysis of FSES 3++ on EGSD 24.00.00 "Aviation and Rocket-Space Engineering" and professional standards in the field of Rocket and Space Industry and Aircraft Engineering was performed in terms of the necessity to include the study of AI methods in professional training.
2. The courses and individual modules on the basics of AI and data mining methods were developed for implementation in the curriculum of the EGSD 24.00.00.
3. Recommendations on the implementation of both general courses and individual modules of AI methods into the training system of specialists have been prepared.

3 Analysis of FSES and professional standards for EGSD 24.00.00, and preparation of prospective courses on AI methods
Analysis of the FSES (Federal state educational standards) and professional standards combined with the AI methods and data mining was conducted to determine the focus of the individual modules and general courses development.Seven FSES of 3++ generation were studied, and the majors for which it is feasible to include both individual modules and comprehensive training on AI methods were determined.Integrated training is recommended for the training programs 24.05.03 (Aircraft testing); 24.05.04 (Navigation and flight support of space vehicles); 24.05.05 (Integrated aircraft systems); 24.05.06 (Aircraft control systems).

Table 1 .
Modules and contents of the general course on AI methods NN: VGG, GoogleNet, LeNet.Reinforcement learning.Functions for describing agent's behavior.Deep Q Learning algorithm.Open AI Gym library.Basic features of the Neural Toolbox package of Matlab

Table 2 .
DATA Mining course syllabus