Creativity in Intelligent Technologies and Data Science

The paper dwells on the problem of holistic perception of heterogeneous information by a computing system - text, images, formulas, etc. that remains open. It is to the problem of holistic perception that the task of synthesizing the CPS architecture for implementing adaptive control mechanisms can be attributed. This paper discusses experiments on the synthesis of CPS architectures using open pre-trained generative models, on retraining of a neural network model Stable Diffusion, as well as a new model based on the generative-adversarial approach for the synthesis of CPS architectures.

An effective tool for the formation of intelligent transport systems is machine learning (ML). But applying increasingly sophisticated machine learning technologies is creative. This allows us to speak about the significant subjectivity of the conclusions of ML. Thus, a gap is formed with the theory and practice of management based on ML. Intelligent technologies for managing transport systems using understandable machine learning are called upon to fill this gap. The article discusses an approach to the development of such a technology for a three-level service management system in a corporation. At its top level is the Boss, at the middle level is the Curator, at the bottom level is the Manager. The Boss should increase the scope of services provided by the corporation. But the Curator knows his abilities better than the Boss. In turn, the Manager knows his potential better than the Curator. Thus, both the Curator and the Manager can manipulate the scope of services they provide in order to get more incentives. To avoid this, a service management system is proposed that includes two explainable ML procedures. Sufficient optimality conditions for this control system are found. In their implementation, both the Curator and the Manager are interested in maximizing the scope of services provided. Such a management system provides algorithmic accountability, responsibility, trust and recognition of ML in the corporate team. The proposed approach is illustrated by the example of service maintenance repair of rolling stock.

Building of all spheres of life at a qualitatively new technological level and possession of one's own technological keys to the creation of goods and services of the next generations is necessary to ensure one of the key principles of the development of the state, namely the achievement of technological sovereignty. In modern realities, the development of enterprises cannot be carried out without coordination with partners from Russia, as well as China, India and other countries. The selection of potential partners can be carried out on the basis of the revealed significance of their patented technological solutions. Further ranking of potential partners can be carried out on the basis of the revealed significance of their patented technological solutions. At the same time, it is proposed to use three criteria: the mass nature of the subject of the patented invention in the current period, the predicted mass nature of the subject (technology) in the future period, the success of the patent in the information field. The novelty of the developed method of forecasting the significance of patented technologies is the use of the generated metrics of innovation potential (prospects) to analyze the global patent array according to the sphere of interests of key enterprises of the Volgograd region. The developed software module provides the following functions: a) parsing of patent documents is carried out from Yandex Patents and Google Patents; b) the formation of a list of IPC classes corresponding to the spheres of interests of enterprises of the Volgograd region, and the extraction of patents of these classes from Google Patents; c) the determination of the mass content of the subject of the invention in the current period is carried out by clustering the patent array based on the lists of keywords provided by Google Patents; d) the predicted mass content of the subject (technology) in the future the period is determined by the ARIMA method; e) success in the information area is determined based on the information provided by Google Patents about the citation of the patent.

The paper describes the stages of developing a deep learning model- based method for identifying the pharmacological activity of a synthesized chemical compound. The implemented software is designed to prepare data for training, and testing, using a deep-learning neural network MPNN, obtaining the results of the neural network in the form of a concentration coefficient of half-maximal inhibition. The approaches and technologies used to solve the problems of predicting the activity of a synthesized substance are disclosed.

The article discusses the theoretical foundations of hybrid cyber-physical systems as complex systems in the context of the digitalization of the economy. The relevance of the study is due to the fact that there is a rapid increase in the use of intelligent technologies in all areas of both the real sector of the economy and in the financial sector. The scientific novelty lies in the fact that, in contrast to the previously used methods of algorithmic exchange trading, a new one has been proposed that is fundamentally different from those used previously. It combines both a system for sending orders, functioning on the basis of logical algorithms, and an intelligent system based on machine learning “Random Forest”, which forms a forecast of the closing price of a financial instrument, the joint operation of which, in the process of responding to market changes, ensures a decision on buying/selling in automatic stand-alone mode, thus enabling efficient speculative “intraday” trading. The practical significance lies in the fact that the application of this development provides an increase in competitiveness in exchange trading due to highly profitable speculative operations on an hourly timeframe based on an accurate forecast. The implemented software has a Certificate of Rospatent for a computer program No. 2022662398 dated 06/22/2022.

This article is dedicated to a new method for translating audio recordings of interjections between two domains. The original speaker's voice is transformed to the target one within a given voice pair, and vice versa. Basing on an overview conducted, mel-frequency cepstral coefficients are selected as main features of signal. The introduced method is implemented and approbated on the grounds of a book-reading training dataset and sample interjections for a number of voice pairs, including human-robotic. Recommendations are given on the method applicability and on dataset recording. Obtained results testify that we found the solution to the problem of overcoming constraints of existing speech-synthesizing software, namely that of the limitedness of interjections forms and that of the poor intonations variety. The method is to be applied in order to fill reactions databases for dialogue systems designed for affective communication, such as the F-2 interlocutor robot. The introduced method will enable spoken reactions corpora developers to record interjections of interest and to translate them to the selected synthetic voice.

This article is dedicated to automatic detection of natural language texts functional style. Part-of-speech N-grams are selected as text features for capturing word order, which depends on functional style in Russian. The introduced approach was approbated within the task of texts classification by functional style and within a content-oriented book recommender system, which uses basic and modified probabilistic topic modeling and selects writings basing on their styles similarity to the input. Successful style-based books selection showed that an expectation gap for recommender systems can be filled, since it became possible to match books by style, successfully crossing the genre boundaries. The results are applicable for texts classification by functional style in libraries and publishers software, for personalized writings selection in recommender systems for news, scientific articles and fiction, as well as for further style modelling in dialogue systems and conversational communicative robots in order to select appropriate style depending on the interlocutor’s one and on the polilogue pragmatics and participants’ roles.

Modern visualization methods are used to convey information about an object or process and as a tool for search and decision-making process. Data and signals, in analog and digital form, are only valuable if they are analyzed for a specific goal. In this work we etablish the classification of visualization tasks from the point of analyzing heterogeneous multidimensional data, including the case when at the initial stage it is required to formulate a research hypothesis. A classification of visualization metaphors is presented, which is necessary for a conscious choice of tools for visualization and data analysis. This is important for understanding and managing the interpretability of information, the formation of the correct meaning and operational understanding. We demonstrate examples of static and dynamic models of visualization. Based on the semantic model and proposed classification, the principles of visual metaphors formation for solving several applied tasks in various fields of knowledge (oil and gas production, biomedicine, materials science, education, management, etc.) are formulated.

Continuum robots are robots with high flexibly and maneuverability, which allows use them in confined workspaces with many obstacles. Continuum robot’s motion planning and control are depends on inverse kinematics. Existing inverse kinematics solvers have high computational cost and often fail to find a solution. Moreover inverse kinematics solutions for continuum robots with variable section length underrepresented as well as solutions for multisection robots with mixed sections. This paper presents a further development of FABRIK-based inverse kinematics algorithm that allows operating? with multisection continuum robots with variable length. The paper presents analytical solution single section with variable length as well. Our experiments show that proposed algorithm show have higher solution rate and lower solution time in comparison with Jacobian-based inverse kinematics.

The authors dwell on the problem of appointing rational cutting modes for machining steels and alloys with a coated carbide tool. To solve it due to operational information about properties of each “coated tool–workpiece” contact pair, the authors have proposed to use the thermoEMF signal value (mV) of a dynamic thermocouple recorded during a test run. To justify the practical relevance of the method proposed, the theoretical foundations and study results of the thermoEMF signal information capacity when evaluating physicomechanical properties of the “steel billet–coated carbide tool” contact pairs for setting rational cutting modes are presented.

The relevance of the article is justified by the increase in the speed and intensity of traffic in Russia and in the world, and, as a result, a decrease in the distance between cars in the traffic flow, leading to an increase in the frequency of accidents. To reduce the number of such situations, modern cars are equipped with an anti-lock braking system, the effectiveness of which is determined by the algorithms implemented in it and its control system and interaction between the controller and actuators. The scientific novelty of the research lies in the combined approach to the study of the use of fuzzy logic elements in the electronic control unit of the braking system. The study of the possibility of using a fuzzy controller to calculate the braking parameters of an automobile wheel equipped with an anti-lock system for different types of road surface has been carried out.

Technological maps are the primary planning document and economic analysis in an agricultural enterprise and its divisions, serve as the basis for the development and adoption of specific management decisions in the crop production industry, financial and long-term plans of the enterprise. Technological map planning is currently carried out by compilation using various tools or in paper form, and can take up to several working weeks, depending on the land bank of the enterprise and the depth of planning. This paper describes the developed method for analyzing the costs of performing the stage of field work within the framework of the technological map of cultivation of an agricultural crop, based on the processing of statistical data on the results of the actual performance of field work. The developed method has been tested at an enterprise with an average field area in Russia (6,500 hectares). After the introduction of a management system at the enterprise with a planning tool for technological maps of growing crops, the time spent on planning was reduced by 50%. When using the proposed method, this indicator improved by another 20%.

A high-precision forecast of the hydrological regime of a large river system is limited by poor knowledge of the dependence of the hydraulic resistance force on the flow parameters. Our research is aimed at developing a method for constructing a hydraulic resistance model based on the integration of direct hydrodynamic simulations and intelligent computing. The force of hydraulic resistance to flow is the sum of the impact of roughness, traditionally given by the Manning coefficient, and the force of turbulent friction. There are four free parameters in our hydrodynamic model, which are calculated by fitting the results of computational experiments with data from real measurements at gauging stations. Our analysis focuses on the hydrological regime of the Volga River in its lower reaches. We apply the Long Short-Term Memory neural network to determine these free parameters of the numerical hydrodynamic model, requiring the best agreement between the measured and model time series of water levels at three gauging stations during 2022. The dependences of water levels on the water discharge (hydrograph) in the channel show the presence of memory in the system, when the value of the water level depends on the behavior of the hydrograph over the previous few days. This leads to the appearance of a hysteresis-type dependence in the hydrological data for all three gauging stations. We define the structure of our training subsets, which allows us to determine the duration of such hydrological memory in the conditions of the Lower Volga. The best agreement between the time series of measured and model water levels is achieved for time interval of 4–6 days. The constructed solutions make it possible to qualitatively reproduce the ambiguous loop-like functions of the water level on the hydrograph of the hydroelectric dam.

Within the framework of the new concept of strategic management and entrepreneurship, it has become customary to represent an organization in the form of a business model reflecting core aspects of the organization in economic, social, cultural or other contexts. The article is devoted to the development of a concept of constructing business models, reflecting new ways of existence and functioning of organizations in conditions of continuous interaction with the information environment. A model of an organization as a fragment of an information network formed by a set of open interacting organizations has been built. The state of the system in terms of the finite network is modeled as the result of its interaction with the environment, determined by the input and output information flows. Changes of the input and output flows leads to changes in the system state. A topological model of a digital information space is built, which is represented by interconnections between organizations through input and output information streams. It is shown that in the space there always exist contours formed by chains of information flows, connecting organizations with each other. A method has been developed for finding the contours that determine the mission of each organization in society. The following concepts are proposed as characteristics of the contour that determine the current state of the organization: the length of the contour, the strength of the connectivity of the contour, the potential of the contour. The principle of building business models are based on considering the of managing the organization is formulated as the task of ensuring that the system belongs to the contours formed by the chains of information flows that ensure the inclusion of the system in the information space, predicting the emergence of new needs and inclusion in new contours to meet these needs. This approach allows us to correlate local goals and goals of society and organize management in organizations from the point of view of cooperation of goals. As part of further research, it is planned to develop a management support system (MSS) for supporting planning and decision making in organizations based on the proposed models and methods.

In this article, the methodological basis of decision support for controlling the process of providing services is designed. This framework includes new management mechanisms based on incorporating the analysis of personal and emotional information into the decision support process. This allows us to solve problems with poorly formalized subjective information, create models that reflect the properties of real objects, and increase customer satisfaction. The models and decision support methods developed formed the basis of a prototype decision support system (DSS).

The analysis of the dynamics of economic processes and, in particular, innovation processes, has a great significance. However, the study of dynamics in economics is complicated by frequent qualitative changes in the time series. The problem is also complicated by a small amount of data, as they are often generated on the basis of annual reports. In this paper we propose to use an algorithm developed by the authors as an instrument of analysis and modeling of the socio-economic processes dynamics. This algorithm is effective taking into account the features of these processes. Our algorithm consists of preliminary decomposition of the time series into components by the SSA method, modeling of these components and selecting the best model basing on fuzzy rules. In this paper, the algorithm is used to study the dynamics of the revenue from new (innovative) products and services in some regions. As a result of the study, dynamics models were obtained that give a reasonable fit to real data and also give realistic forecasts. It has been shown that our model can be used for analytical purposes. The results of the work may be used in the innovation activity analysis in the regions and should be taken into account when making decisions in the management in the socio-economic sphere. The created algorithm proved to be effective for analyzing and forecasting of the innovation processes dynamics. The peculiarity of the algorithm is that it greatly simplifies the comparative analysis of the dynamics series.

The article describes the development of methods and technologies of production planning as part of a scheduling for large-capacity multi-assortment production. The article presents a systematic analysis of this class of production characteristics and also proposes a formalized generalized statement of the mathematical problem of a production planning. The proposed methods and technologies can significantly reduce the production planning time for large-scale production planning tasks, as well as speed up the possibility of its rapid adjustment. The proposed software package allows reducing the time of production and simplify the process of making managerial decisions in the implementation of production planning.

The intellectual technologies of computer expert systems are presented, which allow designing various chemical and technological facilities, including with the help of digital information models. The information support of these systems is described, including updated and supplemented knowledge bases developed on the basis of intelligence maps that allow graphically structuring information about the main characteristics of a production facility. Examples of the functional structure of intelligent systems for the processes of secondary oil refining and processing of industrial waste are given, increasing the efficiency of design and planning.

This paper presents findings of the research proving that the known method for automated assignment of rational cutting modes, in unmanned technology, to modern CNC lathe machines, which are essentially cyber-physical systems, can be applied to processing with the use of lubricating and cooling technological fluids. This requires both conducting a certain number of experiments and building appropriate regression mathematical models. These studies are aimed at teaching cyber-physical systems in mechanical engineering to make independent decisions in assigning machining modes based on preliminary or operational information on the thermo-physical properties of contact pairs “tool-work piece” and other measuring signals from the cutting zone without operator participation.

The subject of the research is the approaches and methods of digital processing of images of slices of meat products related to the determination of the contour of the selected fragment, its shape, area and colour. The goal is to develop an algorithm for the automated detection of falsification in meat products based on the processing of digital images of histological sections. As a result of the study, a comprehensive analysis of the process of detecting falsification in meat products was carried out, and specific tasks requiring automation were formulated. The following tasks have been completed: determining the contours of the counterfeit image, determining the area of the regions and the shape of the particles, determining the colour of the counterfeit particles. To develop the developed algorithm of four modules, the Spark Streaming interface and the Python programming language were used. The architecture of a hybrid expert system, previously proposed by the authors, was used to support decision-making when determining the presence of a falsification. The developed software system was tested. After executing the four Spark Streaming modules and displaying the results, the user enters data into the decision support system: particle size (area), shape, draw a conclusion on a hypothetical topological invariant, select a colour from the drop-down list. The system automatically formulates a conclusion and gives the name of the counterfeit. If there is no falsification on the image, the object detection module does not allow it to be further processed by the system. The area of application of the results of the work is the production enterprises of the food industry with automation of the processes of quality control of meat products and the detection of falsification; regulatory organizations and supervisory bodies to improve the effectiveness of quality control; laboratories and scientific research centres for automatic analysis and image processing.

The research is aimed at building a digital workflow (DW) that implements an iterative procedure to create high-quality dental products for individual patient needs based on cobalt-chromium metal 3D printing techniques. The introduction of an additive technology for selective laser alloying of metal powder provides a reliable solution to this problem without expensive auxiliary operations. An important component of our digital workflow is the numerical simulation of temperature dynamics during the manufacture of dental metal structures, which affects the strength properties of products. Strong heating and cooling during metal 3D printing cause mechanical deformation due to spatial temperature inhomogeneity. The results of numerical simulations make it possible to mitigate this negative factor by adding complementary temporary support structures directly at the 3D printing stage.

The purpose of this work is to improve the efficiency of monitoring the composition of wastewater by developing a method for detecting anomalies in time series. The paper describes the concept of the system for collecting data from the pH/ORP sensor of the hardware-software complex for automated control of the composition of wastewater (PAK), the proposed approach for the automatic detection of discharges in wastewater using binary classification, as well as the method of automating the process of sampling and retraining the model based on the results of identification, taking into account the expert opinions of specialists. The result of the work is developing a more efficient method for finding anomalies in a time series using binary classification.

The paper proves the necessity to optimize the time of a traffic light signal as the fastest and most cost-effective way to improve the efficiency of the urban street and road network in a continuously growing number of vehicles. The main traffic management strategies are analyzed and the conclusion is made that in some cases it is advisable to have effective software that switches the operation modes of traffic lights during peak load periods. The aim of the study is to develop a method of traffic management based on the application of a fixed time control method, taking into account the uneven traffic intensity and structure of the traffic flow during the day. The authors chose a segment of the road network in the city of Volgograd as an example of its use. The authors suggest a method for software calculation for traffic light objects in a section of the road network, minimizing the loss of time for vehicles in the morning and evening periods of peak traffic. The technique is based on a neural network with an error back propagation algorithm that uses data on adaptive traffic light cycles from SUMO package for neural network training. In the conclusion, a comparison is made about the efficiency of traffic organization on the urban street and road network under analysis with the existing traffic organization and with the developed plan for switching signals of traffic lights.

Rubber-metal shock absorbers are the most common type of shock absorbers used in modern machine-building structures for various purposes. The basis of such shock absorbers is a combination of a technical rubber-like material and a metal shell. The modern nonlinear theory of elasticity, at the moment, cannot provide a generalized equation of state of rubber-like materials, similar to Hooke’s law in linear theory. The development of new materials leads to the need to create new mathematical expressions for the deformation energy potentials. Existing commercial packages allow you to calculate a small number of hard-coded deformation models. In this paper, a specialized calculation system is proposed that allows automating the calculation of rubber-metal shear shock absorbers under finite deformations, freely introducing any generalized neo-Hook deformation energy potential and a rather complex configuration of the shock absorber cross section. The results of calculations of the Treloar, Fan and Gent-Thomas potentials are presented. An estimate of time costs has been made.

The paper presents a digital platform for social analysis using the data retrieved from social media. Based on analysis of multiple parameters used to describe social satisfaction and well-being and regional practice of their integrated monitoring for decision-making support there was proposed single indicator called the “level of positivity”. This indicator was calculated on the basis of semantic and statistical analysis of time series describing the dynamics of social media members’ sentiment changes and estimated using an artificial neural network. Analysis results are visualized by a digital integrated platform in the form of thematic widgets and used to propose countermeasures. Additional attention is given to dynamic analysis of the positivity level deviations in time using approximation models for individuals and groups of social media members. The introduced platform was implemented in practice of the regional ministry of social protection and support and used to identify individuals and groups of people with depressive and deviant behavior in order to conduct timely prevention of negative trends increase the effectiveness of social assistance activities.

The article discusses the model and method of decentralized secure storage of digital portfolios of students in an open educational environment based on distributed ledger technology. The issues of the formation of a digital portfolio, its composition and structure in the process of continuous training of specialists are considered. The portfolio is a digital trace of educational and professional activities that captures the qualifications and emerging competencies of a future specialist for potential employers and recruitment agencies, which increases the competitiveness of graduates of educational institutions in the regional labor markets. The model of a set of digital portfolios of students in the information space is presented as a set of acyclic digraphs grouped by areas of study, educational programs, groups of competencies being formed, etc. To organize storage and work with digital portfolios, it is proposed to use the following technologies: a distributed registry, hashing, smart contracts, decentralized applications, an interplanetary file system. The main area of research is the creation of a multifunctional platform with smart contracts for the transition from a distributed data registry to a digital portfolio management system with personal data protection and support for the business logic of the educational process of forming professional competencies in the training of specialists in an educational institution. The Ethereum blockchain is used as the base platform. Transactions are implemented by smart contracts that perform the operations of collecting, verifying and authenticating data from sections of the information space associated with the student and adding information to the portfolio. Data processing in the distributed registry of digital portfolios is implemented at three functional levels. The digital portfolio includes public and private parts, which are accessed by multihash.

The article considers the issues of planning and managing a tourist product based on the synthesis of the optimal route of movement and visiting tourist sites. A hybrid approach has been developed to manage the process of synthesis of optimal tourist routes. The developed model and method of tourism product management allows taking into account tourist preferences and restrictions along with various external factors that affect the choice of places to visit and route sections. The process of synthesis of alternative tourist routes is carried out at the first stage by a modified ant colony algorithm. At the second stage, the method of analysis of hierarchies was used to select the optimal route from alternatives. The advantage of the approach is the ability to control the route synthesis process in two ways. The first method takes into account the importance of the experience, preferences and limitations of a particular tourist when choosing sites and tourist sites on the route to visit at a particular point in time. The second way is the ability to manage the workload and the number of tourists on route sections and tourist sites, as well as the attractiveness of objects for tourists by changing the degree of importance of the concentration of pheromones when calculating the probability of choosing a site or visiting an object. The process of planning, designing and operating tourist products with the ability to control the process of selecting objects on routes is implemented in a recommender tourist system.

Adaptive learning applications that provide personalization of the learning process are one of the fastest growing types of educational software. The article analyzes the models and methods of adaptation for the development of adaptive learning games in which the learning process is implemented in a game context. A portable adaptation model is proposed, applicable for the development of adaptive learning games of various genres. The learning course for adaptive learning games is represented by a space, formed by embedding the initial course structure into the lattice, the structural ordering of which allows the formation of personalized learning strategies determined by the current state of the learner and the structure of the space itself. An approach for dynamic content matching of non-linear learning scenarios and gameplay is proposed, which preserves the logic of both the learning process and the game process. An embedded adaptation module for adaptive learning games has been designed and implemented, which makes it possible to reduce the complexity of development and improve the quality of such games in terms of the effectiveness of the learning process. An example of applying the proposed module to the development of an adaptive game for studying the object-oriented paradigm and the C# language is described.

The paper examines the possibility of using companion robots (F-2 robot, as the example) to assist humans in learning the tonal system of Chinese. The experiment compares the effectiveness of supportive cues via speech and gestures. The results of the experiment (N = 20, 4 males, mean age 24,5) show that the learning is effective – subjects learn the pronunciation of Chinese syllables with the given tones. At the same time the effectiveness of speech cues is higher than of the gestural cues. It is the speech cue that seems to be more understandable during the subjects pass the introduction to the system of Chinese tones. In general, a combination of conditions in the experiment is effective when subjects learn Chinese phonetics with gestural cues and with speech cues – in any order of conditions. In addition, the robot with gesture cues is perceived by the participants as more natural and engaging, eliciting more sympathy. The results suggest that companion robots with gestural behavior can be used to support the educational process and increase students’ engagement.

The paper investigates the effectiveness of methods of interaction with virtual objects in a virtual reality environment. We consider the role of various input devices, such as virtual reality gloves and vests, in providing a more realistic and immersive user experience, analyze the advantages and applications of these devices in various industries, including medicine, education, the gaming industry and others. We also describe the development of a VR simulator for modeling emergency scenarios, consider the implementation of realistic physics of objects in the virtual world, taking into account the use of Configurable Joint and Rigidbody.

In Mongolian National University of Education sometime need to teach a single teacher several courses in parallel during one academic semester. In this paper describes the evaluation process and data analyze of the five different courses teach by the single teacher. For evaluation process is applied the structure-oriented evaluation model. 69 students responded to survey checklist. All evaluation scores were higher than 0.73 but less than 0.87 which confirmed prediction of evaluator teacher. The teacher is predicted that evaluation results cannot but enough high due to the complex of teaching many different courses in parallel. Main aim of this study was figure out opinions of students and compare the evaluation result with prediction of the teacher. Based on this fact make in light some teaching issues to stakeholders.

The scanning window with the calculation of the local two-dimensional Walsh transform in it is proposed to be used for automated systems of processing ultrasound images. Preliminary studies of the computational capacity of such a process have been carried out and the high speed of this approach has been shown. Frequencies are selected on the spectral two-dimensional plane. The analysis of these frequencies makes it possible to determine the degree of echogenicity in the scan area, as well as to detect the edges of ultrasound image objects. The architecture of the ultrasonic image pixel state detector is proposed based on a two-dimensional scanning windowed Walsh transform, a significant frequency selector, and a neural network tuned to classify the pixel state. The detector was tested both on simulated images and on ultrasound images of phantoms and real medical ultrasound examinations of the human abdominal cavity.

One of the trends in the development of modern mobile applications is adaptability, which makes it possible to increase the accessibility of mobile applications for various categories of users. The article is devoted to the study of approaches to the development of adaptable mobile applications and software development tools for such applications. The possibilities of using CALS-technologies to support various stages of the MA life cycle are considered. A technology for the development of MA, the interface of which can be adapted at the maintenance stage, and ways of adapting the interfaces of the developed MA for the end users are proposed. A system has been developed that combines the typical capabilities of mobile application management systems at the stages of implementation and maintenance, and expands them by adding functions that provide the ability to adapt the MA user interface. MA development is based on a multi-module architecture, which makes it possible to use MA components for reuse in other MAs. A template-based technology for adapting the MA interface and a method for adapting the interface using a built-in configuration panel (CP) have been developed. A method of using the system for developing adaptable MA is described. It is shown that the proposed technology makes it possible to increase accessibility of MA and applicability for different people with intellectual disabilities by customizing the interface for each user in accordance with his/her capabilities and limitations and significantly reduce the overall development time of MA through the use of a unified multi-module architecture and the inclusion into the MA project of ready-made modules, as well as the generation of CPs for adapting the MA interface.

The article discusses the concept of the human state, its types, methods of assessment. The state is a reaction of the body and psyche to external influences. The human state is the most important part of the entire mental regulation, play an essential role in any kind of activity and behavior. Considered types of functional states that affect human activity: fatigue; monotony; mental overstrain; tension/stress. The study presents the main technologies that can be used to assess a human state such as oculography and emotion recognition, the input parameters used for this and the ways to collect them. The main approaches of emotion recognition are considered, such as emotion recognition based on facial expressions; speech-based emotion recognition; emotion recognition based on physiological signals; gesture-based emotion recognition. As well as the main approaches of oculography: electrooculography; videooculography.

In urban passenger transportation (bus or trolleybus), the most common traffic situations are exit from the traffic flow and embedding in it, as well as the passage of regulated and non-regulated intersections. It is on the clear and skillful actions of the driver in these traffic situations that the safety in urban passenger transport mainly depends. In connection with the responsibility for the lives of passengers and other road users, the work of a driver is associated with high neuro-emotional stress. The intensity of the work of the driver of urban passenger transport is explained by the complexity of driving a vehicle in conditions of heavy traffic, shift work schedule, regulated time and speed along the route, and other factors. The high level of functional readiness of the driver before the trip is an important component of safety in the implementation of urban passenger transportation. In order to increase the prognostic significance of this approach, the key factors that reduce the level of pre-trip performance and significantly affect the level of the driver’s functional readiness are identified. Simple and publicly available methods for assessing the functional state have been developed and selected, original and existing test tasks have been used to diagnose the degree of development of professionally important qualities of a driver. Software for personal computers has been developed, which makes it possible to carry out express diagnostics of the functional states of drivers based on the analysis of the obtained test values.

Nowadays, sleepiness research is particularly relevant in fields like transportation, where drowsiness can lead to accidents and fatalities. Developing an accurate sleepiness detector requires a deep and fundamental understanding of sleepiness processes and mechanisms. The purpose of the current study was to investigate the features of evening-night situational sleepiness and heart rate metrics in individuals with different levels of daytime sleepiness. A collection of 32 recordings was gathered from the Subjective Sleepiness Dynamics Dataset. Daytime sleepiness was assessed using the Epworth Sleepiness Scale, while various domain heart rate variability (HRV) metrics and situational sleepiness (measured by the Karolinska and Stanford Sleepiness Scales) were assessed at 8 PM and 10 PM. The study results demonstrated that situational sleepiness increased from 8 PM until 10 PM only in individuals with lower normal daytime sleepiness, which was accompanied by a decrease in TINN, possibly indicating an increase in fatigue. On the other hand, individuals with higher normal daytime sleepiness did not experience a change in subjective sleepiness, but their sympatho-vagal index decreased, and fragmentation heart rate metrics increased from 8 PM to 10 PM. Thus, the results confirmed the hypotheses regarding a significant increase in subjective sleepiness in individuals with lower daytime sleepiness from evening till night, and the different dynamics of HRV metrics from evening till night in individuals with different daytime sleepiness levels.

The concept of a system for collecting and analyzing data by region is proposed, and as a result, obtaining an integral indicators system that allow a comprehensive assessment of the individual regions state. Possible areas of such system application are considered, ranging from the tasks of general resource planning for optimizing the socio-economic well-being of individual regions, and ending with the tasks of zoning on an administrative-territorial basis. An algorithm for such problems solving is described. The features that must be taken into account for the Russian Federation subjects are listed. Key research methods are identified and it is noted that an important component of the solving this problem process is proactive modeling, which makes it possible to predict some consequences of infrastructure changes, the economy and the social sphere in a particular territory. An important advantage of the described in this paper system is that the data obtained from different sources and fed to the neural network input are naturally integrated with each other, allowing for a comprehensive assessment of the regions state.

The paper investigates the technology of organizing the transportation process in the passenger urban transport system based on unmanned automatic vehicles. Recently, research has been intensively conducted in the field of development of unmanned vehicles in many countries of the world. On the basis of these vehicles, fully automatic transport systems should be developed in the future, excluding the human dispatcher from the control loop of urban passenger transportation. Such systems have parameters and capabilities far superior to modern urban transport systems.

This paper presents a solution for using generative design techniques with ontology engineering to create park layouts based on normative documentation. Generative design is a methodology based on the use of algorithms and applications to generate iterative and variable design solutions. It allows the analysis and consideration of multiple factors such as topography, climatic conditions, human flows, transport networks and other parameters to develop the best solutions for a specific site. Ontology engineering is concerned with the creation of formal descriptions and models to represent knowledge about the domain. In urban planning, ontology engineering can be used to create a formal model of a city that integrates information about its physical environment, infrastructure, public spaces and transport network. Combining generative design and ontological engineering in urban planning opens up new possibilities for designing innovative and optimised urban environments. Generative algorithms can use ontology-based knowledge models to automatically generate and evaluate different urban design solutions, given parameters and objectives. This makes it possible to explore a large number of options and find optimal solutions taking into account multiple factors. The article analyses the use of generative design and ontology engineering technologies in the field of urban planning. The ontological model based on the normative documentation - SP 475.1325800.2020 “Parks. Rules of urban design and landscaping”. A method of creating a park zone layout using generative design with the use of knowledge represented as an ontological model is proposed. An example of the implementation of the proposed method using the cross-platform computer game development environment Unity is given.

Modern cities meet many of the criteria for a comfortable and safe life for their citizens. However, the rapid researches and developments in the fields of information technology and telecommunications stimulates further modernization of urban environment in order to provide maximum possible conveniences to everyone - from a resident to an urban planner and manager. The work of the latter largely determines the comfort and safety of urban residents. In this regard, it is necessary that urban planners and managers can make decisions based on objective factual data, which needs to be easily and quickly obtainable without spending a lot of time on it. For this reason, there is a need to develop IT solutions aimed at the optimization of the work of planners and managers. In order to justify solutions for transforming the urban environment to make it more comfortable, safe and harmonious, sustainable and integrated development, methods for analyzing and assessing the current state and monitoring changes are needed. This paper considers the problem of collecting data on the state of urban infrastructure. The application of spatial data analysis to solve the problems of urban territories management and urban planning are investigated. A method for estimating the area of the city, occupied by the objects of different categories, is presented. Implementation of the proposed approach is shown on the example of calculating the area occupied by industrial objects.