The paper is dedicated to the key information technologies, which are important for the economic progress of various countries. Common properties of the IT sphere are discussed. Examples of research work at Ufa State Aviation Technical University are given for the topics of Big Data, Arti cial Intelligence, Robotics and Sensorics.
eld of information technology Copyright c by the paper's authors. Copying permitted for private and academic purposes. system administrators, maintenance specialists, java programmers and developers, web-programmers, testers, PhP-programmers, deployment specialists, web-designers, technical writers.
Regional issues of IT-development mostly coincide with issues at the federal level. In particular, these are insu cient nancing, personnel shortage, inconsistency of actions between di erent levels of government. However, there are some peculiarities in the regions, e.g., a weak development of the information and telecommunication infrastructure. Another problem is that in Russia in recent years no signi cant bank of solutions or services has been established at the federal level, which regions can use together and at no additional charge; in fact, every region is seeking funds for the same solutions.
Nine key IT for the development of digital economy are stated in [PDE17]. They are listed in table 1.
Processing large amounts of data (large in comparison with standard scenarios); work with fast stream of these data; parallelized work with unstructured and weak-structured data. Arti cial intelligence (including arti cial neural networks) is a technology for creating intelligent machines or programs, which are capable to model human intelligence for implementing complicated task, such as natural language text processors, expert systems, virtual agents, recommendation systems, etc.
independently by each node of the network. The most wellknown type of distributed ledger is blockchain. Security of the stored data is provided by speci c use of cryptographic hash functions and digital signatures.
Manipulation of complex quantum systems at the level of their individual components in order to create powerful and secure quantum computing system.
ulation and modeling, additive manufacturing, alternative energy, industrial big data and analytics, development of cyberphysical systems and "Industry 4.0" concept.
Subcategory of the Internet of Things; a concept for building information and communication infrastructures: connecting any non-residential devices, equipment, sensors, control systems to the Internet and integrating the elements among themselves.
Robotic components intensify industrial production. sensors in robotics act as receptors through which robots receive information from the outside world and their internal organs. Transferring information between points without wired communication
of human life with objects created with digital devices and programs and having the character of an image. Augmented reality allows placing visualized information around the objects of the real world.
Applications medical and socioeconomical data analysis, GIS, etc. medicine; unmanned vehicles; management of human resources, industry; nance; household robots, smart buildings, etc. banking, nances, energy systems, etc. cryptography, arti cial intelligence, molecular modeling, etc.
Industrial production
Industrial Internet, Internet of Things, communication systems informational and educational products; games; product development; presentation and demonstration
The scientists from the Faculty of Computer Science and Robotics at Ufa State Aviation Technical University made signi cant work in the eld of fundamental research and in the eld of practical research (in collaboration with customers from the real sector of the economy). Several tens of examples of research projects connected to key IT may be given. However, due to the limited scope of this article, we will focus only on some results in the eld of big data, arti cial intelligence, robotics, and sensorics.
Big data analysis in Geo-technologies are now at thbeing developed at the Department of Geo-Information Systems (Prof. O. Christodoulo). Geo-information systems (GIS) are becoming one of the main technologies for analyzing big data for many applications, such as production management, transport, energy, security, etc. Geodata are complex, and with the transition to big geo-data complexity increases almost exponentially. It means processing billions of geographic objects, (lines and polygons). In addition, the calculation of spatial relationships is often required. Applying technologies like Hadoop to them could help to solve problems that cannot be solved by traditional methods, or, at least, to get a signi cant gain in e ciency and speed of computations. The following research works in processing big geo-data were conducted: development of GIS of trunk oil pipelines and 3D models of potentially hazardous objects for OJSC "Uralsibnefteprovod", development of an automated system for the formation and maintenance of a regional waste cadastre in the Republic of Bashkortostan, development and implementation of the GIS for OJSC "Gaz-Service", development of the GIS for governing organizations in the Republic of Bashkortostan, modeling the ood zones during the period of spring oods in the Republic of Bashkortostan [Chr13].
Research work on BigData at the Department of Computing Technology and Information Security is made together with Frodex LLC on the research topic: Intelligent analysis of banking transaction data as part of a antifraud system [Sap17] (Prof. V. Vasilyev). Within the framework of the project, the system for collecting and processing user data was developed as part of the antifraud system. The key element of this system is the data mining module. Analysis algorithms are applicable in the conditions of big data. Hadoop hardware and software cluster is deployed. It consist of 16 server machines that implements the capabilities of a distributed transaction banking data processing system based on modern distributed le system technologies, non-relational DBMS (Cassandra), big data processing technologies (Apache Spark,TensorFlow, Spark-sklearn).
Research on medical applications of Big Data are mainly conducted at the Department of Computational Mathematics and Cybernetics (Prof. N. Yusupova, Prof. G. Shakhmametova). The most relevant direction of work is data processing for diagnosis, treatment and prevention of diseases. Taking into account high volume of input information or implementing a complex data processing algorithm may be really di cult for a decision maker. Relevance of this task increases with the accumulation of information, which is happening avalanche-like due to the improvement of information collection and storage technologies. Tasks of medical analytics solved via BigData technology may be divided into descriptive analytics (What happened?), diagnostic analytics (Why did it happen?), predictive analytics (What will happen?), and preventive analytics (What should be done in order NOT to happen?). Medical data are obtained from various sources:results of research and testing; medical records and diagnoses; medical meters; analytical information from medical authorities and pharmacies.
In collaboration with the Department of Internal Medicine Propsy at the Bashkir State Medical University (Prof. R. Zulkarneev) a technique for complex analysis of toxicologic data was developed and implemented [Sha18]. This technique includes three main stages: exploratory data analysis using visual analysits, nonparametric analysis, data mining. The developed technique is applied to the analysis of data on acute poisoning in the Republic of Bashkortostan in 2015-2016. The results obtained during the analysis can assist the managers of medical institutions and chief specialists of the governing bodies in analyzing the indicators characterizing the dynamics of trends in public health, planning the allocation of health care resources in the region, and managing specialized medical services. Additionally software complex for non-parametric medical data analysis methods was developed. Future plans include descriptive and diagnostic analytics of medical data; study of the e ectiveness o f treatment based on analysis of medical records and diagnoses; expense forecasting based on the analysis of data such as the number of repeated visits, the prevalence of pathologies, the number of patients with chronic diseases, etc.
Intelligent analysis of big social and economic data is also performed at the Department of Computational Mathematics and Cybernetics (Prof. N. Yusupova, Prof. O. Smetanina). Solutions are based on text analysis, pattern recognition, machine learning [Guz14, Yus18]. Main instruments are data quality assessment, factor analysis, dimensionality reduction, cluster analysis. One example task is odenti cation of similar educational programs in di erent countries with preliminary processing of semi-structured and structured data. Another one is identi cation of hidden patterns in personal data, including the data from Internet-resources. Results are used for the selection of job-applicants, for assessing the loyalty of customers (including students as customers of educational services), for assessing the correct choice of a professional path, etc. Analysis of natural language text was applied for review-based decision making. Pattern recognition as a mix of classi cation and identi ca tion methods was applied for person recognition based on voice and images.
Research in the sphere of intelligent robotics and sensorics is conducted at the Department of Computer Technology and Information Security (Dr. Konstantin Mironov). This work is dedicated to the task of object transportation by robotic throwing and catching [Gay17]. This is a novel possible way of material transportation, which may replace traditional conveyor belts for transporting small objects in the industrial environment. Catching is based on tracking and forecasting the trajectory of the thrown object. Proposed tracking system is based on stereo vision. Novel approach to forecasting the trajectory of the thrown body was proposed. Traditional approach to this task consist in mathematical modeling of the ballistic trajectory. Learning-based approach is less popular. Proposed predictor was applied based on nearest neighbor regression, which does not require exact physical model of the motion. These two approaches were then mixed in the new one, which is based on genetic programmi ng. The forecast is made using an equation, which is learned by the procedure of genetic programming. Results of numerical and throwing experiments showed that accuracy of trajectory forecasting is enough for successful robotic catching. 4
Global IT industry has great growth potential in the long term with a tendency to increase the share of IT services in comparison with software and hardware. The key information technologies of the digital economy include: big data, neurotechnologies and arti cial intelligence, distributed ledger systems, quantum technologies, new production technologies, industrial Internet, components of robotics and sensorics, wireless communication technologies, virtual and augmented reality. Research experience at the Faculty of Computer Science and Robotics (USATU) has a direct relationship to the key areas of information technology, named in o cial documents. Further research is related to both fundamental problems and the application of results for real-world problems. [EU16]
European Commission sets out path to digitize European Industry. Press Release, Brussels, 2016. [PDE17] Program Digital Economy of Russian Federation. Order of the Russian Government #1632-p from 28.07.2017 (in Russian)
M. U. Sapozhnikova, A. V. Nikonov, A. M. Vul n, M. M. Gayanova, K. V. Mironov, D. V. Kurennov Anti-fraud system on the basis of data mining technologies. 2017 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), Bilbao, Spain, pp. 243 248