Technical University of Varna

Climate variability amplifies temporal heterogeneity in crop production, challenging uniform varietal recommendations and highlighting the need to integrate genotype × environment interactions. This study evaluated the yield performance and stability of sixteen triticale (×Triticosecale Wittmack) genotypes over three consecutive growing seasons (2022/2023, 2023/2024, 2024/2025) at a single location with pronounced interannual climatic variability. Grain yield ranged from 3.49 to 6.68 t/ha in the least productive season (2022/2023) and from 7.71 to 9.92 t/ha in the most favorable season (2024/2025), with overall genotype means varying between 6.67 and 8.12 t/ha. Stability was assessed using regression-based parameters (regression coefficient and variance of deviations from regression), Shukla’s stability variance, and derived indices describing responsiveness (RI), predictability (PI), genetic risk (GRI), stress robustness (SRI), and yield opportunity (YOI). Results revealed substantial genotype × year interaction, with yield strongly dependent on seasonal conditions. Four genotypes combined high mean yield with stable performance and low interaction-related risk, indicating broad adaptability across years. Another four exhibited strong responsiveness to favorable seasons or elevated instability, increasing production risk despite high yield potential. The derived indices enabled risk-oriented genotype profiling, identifying contrasting adaptation strategies. Multivariate AMMI and GGE biplot analyses confirmed these patterns, providing a comprehensive view of interaction structure and stability. This integrated framework translates stability metrics into practical, decision-oriented descriptors, supporting risk-aware genotype selection under variable climates.

Despite existing studies on creativity, examining human resource management practices alongside knowledge management models for constructing creativity remains lacking. This study investigates employee creativity in small and medium enterprises (SMEs) in Indonesia, using data from 508 respondents within a 254-sample frame and employing partial least squares structural equation modeling (PLS-SEM). The results indicate that human resource management practices and technological innovation significantly influence knowledge management and cultivate competitive work environments that foster creativity. The PLS-SEM model confirmed that human resource management practices and technological innovation have a significant direct effect on employee creativity, as well as indirect effects through knowledge management and competitive work environments. Knowledge management and competitive work environment served as double mediators in the mediation mechanism tested in this model. The findings provide practical insights for managers seeking to optimize human resources and technological innovation to enhance knowledge management and create competitive work environments that boost creativity.

Abstract This study examines the possibilities of waste management that are subject to promoting sustainable development principles at the national economic level. The current situation with waste of various hazard classes and its disposal is considered. It was revealed that 16 million tons of unsorted wastes are located in Ukrainian landfills, and on average, these volumes increase by at least 5 thousand tons per year. Mathematical models were obtained to determine Ukraine’s capabilities for exporting sorted waste to obtain funding for the implementation of sustainable development projects. It was revealed that in the period from 2025 to 2029, according to the obtained models, the potential export of secondary raw materials from paper, plastic and cardboard will decrease by 40%, including paper and cardboard by 404.6 thousand tons, plastic by 192.72 thousand tons, and glass by 28.8 thousand tons. This reduction in the potential export of secondary raw materials will be a consequence of the development of the domestic waste recycling system. As a result of the measures taken to sort and recycle waste, it is possible to reduce the amount of waste accumulated at landfills by at least 2914.4 thousand tons per year. It is recommended that the separate waste collection system be improved, a network of sorting stations be formed, and the pre-sale preparation of secondary raw materials be approached more carefully to maximize profits from their sale.

The report examines the application of hybrid intelligence as a combination of artificial intelligence and human expertise for driving sustainable innovation in industrial and energy systems. The study aims to demonstrate the potential of HI to reduce the energy consumption, carbon emissions, and production costs in line with global goals for sustainable development and climate neutrality. The methodology includes a mixed approach: quantitative data analysis (Bulgaria, EU, global), mathematical modelling and a simulation of a 30-day production process. The results show an average 15% reduction in energy, 10% in emissions and 11% in costs, with the effect being most significant for enterprises with large production volumes. Although the practical implementation uses deterministic Excel simulation with percentage calculations, these methods serve as simplified analogues of artificial intelligence techniques. The simulation validates the conceptual assumptions and provides a statistical baseline for the future integration of advanced AI methods such as regression forecasting, neural networks, and optimisation algorithms.

The paper presents the design and structural features of an automated lock assembly machine using a PLC control unit, robotic arms and PC-based supervision system. The implementation of advanced techniques in manufacturing processes is observed in the aspect of combining novel technical means and approaches. The main goal to increase the quality of the assembled details is achieved through the application of modern industrial equipment and modeling concepts.

This study examines the impact of digitalization on logistics processes in Bulgaria and the European Union through the use of artificial intelligence and related technologies. The aim is to assess the degree of digital transformation of small and medium-sized enterprises in the logistics sector and to analyze the relationship between the level of digitalization and the reduction of logistics costs. A linear model was used, in which every 1% increase in digitalization leads to a reduction in costs by approximately 0.5%. The results are compared between Bulgaria and the EU average for the period 2020-2024, and a BG/EU ratio index was introduced, which measures the relative lag of the country.

The study tracks the trends of change of six vegetation indices. Two from RGB (red, green, blue) camera (indices VARI and MPRI) and four from RGN (red, green, near infrared) camera (indices NDVI, EVI2, SAVI and NIRI). It was conducted in the October 2024 - July 2025 season during which there was an atypical late snowfall that led to frosts during the active vegetation period. 4 winter wheat varieties (Annapurna, Tonage, Pibrac, Chavignon), taken at an interval of 7 days, were studied. The photos were analyzed with the ImageJ software. The vegetation indices were generated from single photos of the plots. The results obtained show that the indices from the RGB camera have a maximum during the grading period and their declines reflect stress. The NDVI, EVI2 and SAVI indices directly reflect the differences in the varieties and can be used for their differentiation. NIRI, regardless of the conditions, reaches a maximum during the grading period.

This paper presents a detailed review of foam concrete as a thermal insulation building material, emphasizing its thermophysical performance and potential for sustainable construction. Foam concrete is a lightweight, cellular concrete containing <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{6 0 - 8 5 \%}$</tex> closed pores, resulting in bulk densities of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$300-1200 ~\text{kg} / \mathrm{m}^{3}$</tex>. At densities of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$300-500 ~\text{kg} / \mathrm{m}^{3}$</tex>, its thermal conductivity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(0.08-0.12 ~\mathrm{W} /(\mathrm{m} \cdot \mathrm{K}))$</tex> is significantly lower than that of fired clay bricks <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(0.60-0.90 ~\mathrm{W} /(\mathrm{m} \cdot \mathrm{K}))$</tex> and ordinary concrete <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(1.40-1.60 ~\mathrm{W} /(\mathrm{m} \cdot \mathrm{K}))$</tex>, though higher than mineral wool <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(0.035-0.045 ~\mathrm{W} /(\mathrm{m} \cdot \mathrm{K}))$</tex> and expanded polystyrene (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.030-0.040 \mathrm{W} /(\mathrm{m} \cdot \mathrm{K})$</tex>). The review synthesizes existing research on composition, pore structure, production methods, and the effects of bulk density, porosity, and moisture content on thermal performance. Comparative analysis shows that while foam concrete does not match the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\lambda$</tex>-values of polymeric or fibrous insulation, it offers additional advantages such as Euroclass A1 fire classification, resistance to biological degradation, reduced embodied energy, and potential for incorporating recycled content. Applications in external and internal walls, roof screeds, and renovation projects are discussed, highlighting its ability to provide both structural capacity and thermal insulation. The findings suggest that optimizing pore size distribution, increasing closed-cell content, and improving moisture resistance could further enhance foam concrete's contribution to modern energy efficiency and environmental targets.

This study compares Decision Trees (DT), Random Forests (RF), and hybrid deep learning models on eight years of ordinal university survey data, predicting a key satisfaction question from related items. Using statistical and visual analyses, we evaluate accuracy, Mean Absolute Error (MAE), interpretability, and efficiency. DT provides transparent rules but lower accuracy and higher variance. RF improves performance with more balanced feature importance and greater stability. Hybrids further enhance accuracy and robustness through nonlinear modeling. Overall, DT is suitable for exploratory, interpretable analyses, while RF and hybrids offer superior predictive capability. Findings highlight that ensemble methods outperform DT, capturing multifaceted influences on student satisfaction and supporting targeted refinements in educational surveys and decision-making.

This study examines the role of the Internet as a source of knowledge, information, and social engagement among adults across different stages of maturity, with a particular focus on late adulthood (65 years and older). Contrary to common assumptions about declining cognitive activity in older age, our results demonstrate that older adults show strong cognitive initiative and actively use digital technologies to meet their informational and educational needs. The findings indicate that the Internet has become a primary source of knowledge, preferred over traditional sources such as books, libraries, or personal interactions. Older adults use it to read articles and books in foreign languages, explore places they cannot visit physically, and learn strategies to manage daily tasks and health independently. Furthermore, digital technologies help maintain and even expand social contacts, compensating for reduced physical interactions. Surprisingly, older adults spend more time online daily than younger age groups, reflecting a shift in how they engage with information and social resources. These findings suggest important implications for education, cognitive development, and social support programs for older adults, emphasizing the potential of digital resources to foster lifelong learning and autonomy in late adulthood.