Mladenka Novaković

Machine learning is widely proposed as a basis for smart wastewater treatment plant (WWTP) management, yet reported performance is rarely benchmarked against trivial alternatives. This study uses six years of monitoring from a municipal WWTP in Serbia serving 18,000 population equivalents, comprising 1017 fully measured days. Total nitrogen removal falls from 79.4% outside summer to 67.4% in June to August (p = 4 × 10⁻¹⁸), in five of six years, while influent load, hydraulic flow and phosphorus removal are unchanged, which points to oxygen limitation. Four model families are benchmarked against persistence and a ten-measurement rolling mean on one chronological hold-out period. Predictive skill depends decisively on forecast horizon: at one step ahead nothing beats both baselines, whereas at five steps nitrogen reaches R² of 0.516 against 0.388 and 0.320. Chemical oxygen demand (COD) never beats a rolling mean. A random forest identifies nitrogen exceedance of 15 mg/L with an area under the curve of 0.90. Explanation of that model shows it recognises a breach already under way rather than one beginning, which sets the operating conditions for its use. Evaluating on interpolated rather than measured days raises apparent accuracy from R² of 0.34 to 0.82.

The construction sector is a major consumer of raw materials and a significant source of greenhouse gas emissions, necessitating data-driven approaches to support circular economy implementation and sustainable project management. This study develops an integrated framework combining machine learning-based material stock prediction, carbon footprint assessment, and Environmental, Social, and Governance (ESG) performance evaluation for construction projects. A dataset of 128 residential buildings was compiled from official use-permit documentation. After dimensionality reduction using variance filtering and Spearman correlation analysis, 25 regression algorithms were evaluated to estimate quantities of concrete, reinforcement, and brick products. The K-Nearest Neighbor (KNN) Regressor achieved the best predictive performance, with mean absolute percentage errors of 10.64% for concrete, 10.23% for reinforcement, and 16.05% for brick products. Predicted material quantities were used to calculate CO₂ emissions across materialization, demolition, and disposal phases under linear and circular scenarios. The results indicate that circular economy implementation significantly reduces total emissions, particularly for concrete, with reductions of up to 97% under idealized full-substitution conditions, representing an upper-bound estimate. ESG assessment using the Delphi method identified environmental indicators as the most significant sustainability dimension. The proposed framework enables early-stage emission estimation and supports informed decision-making toward low-carbon and resource-efficient construction practices.

This pilot exploratory study investigates the use of generative artificial intelligence (AI) as support for HSE professionals in the assessment of simulated industrial work situations. Five professionals with industrial HSE experience assessed six scenarios each, resulting in 30 assessments, of which 15 were completed independently and 15 with standardized AI-generated support. The assessments were evaluated in terms of recognition of key HSE aspects, identification of additional information requirements, adequacy and specificity of control measures, unsupported assumptions, final decisions, and confidence in the assessment. Descriptive results showed slightly higher average values with AI support for recognition of key HSE aspects (85.0% vs. 80.0%), identification of additional information requirements (60.0% vs. 53.3%), and adequacy of control measures (75.0% vs. 70.0%). The specificity and applicability of proposed measures remained unchanged, while confidence was slightly higher in the AI-supported condition. Participants also supplemented AI-generated content with additional professional considerations. The findings suggest that generative AI may provide useful complementary support in HSE assessment, while professional experience and expert judgment remain essential for evaluating and contextualizing AI-generated recommendations.

The increasing prevalence of emerging pesticides in aquatic ecosystems poses significant risks to environmental and human health. This study explores the valorization of date seeds—an abundant agro-waste in arid and semi-arid regions—into functional biochar for the adsorption of emerging pesticides from contaminated wastewater. Biochar was synthesized via pyrolysis at 550 °C for 30 min under a nitrogen atmosphere and characterized using BET and FT-IR techniques. The prepared date seed biochar (DSBC) exhibited a high specific surface area of 307.45 m²/g and a well-developed microporous structure conducive to pollutant adsorption. The optimized DSBC achieved maximum adsorption capacities of 28.3 mg/g for carbendazim and 25.7 mg/g for linuron. The removal efficiency exceeded 90% for all pesticides at pH 6–8 and equilibrium was reached within 60 min. Regeneration tests demonstrated that DSBC retained its removal efficiency of 60.3% and 75.5% for carbendazim and linuron, respectively, after tenth cycles, highlighting its reusability and cost-effectiveness. Significant performance potential was demonstrated via the formed biochar regarding stability when exposed to real wastewater composition. Overall, date seed biochar presents a sustainable, low-cost, and efficient solution for mitigating pesticide pollution in wastewater treatment systems.

This work aimed to determine the most relevant organic pollutants, especially pesticide residues, in water samples from the protected area of Lake Zobnatica in the Vojvodina region, Serbia. Sampling campaigns were conducted within different seasons related to different agricultural activities in the vicinity of the protected area. The intensive use of pesticides led to the more frequent detection of specific pesticides such as Terbuthylazine, alpha-Cypermethrin, and Aminomethyl phosphonic acid (AMPA) during the summer, compared to the autumn campaign. In addition to these pesticide residues, other organic pollutants, including phthalate esters, fatty acids, phenols, and aldehydes, were also detected. The characterization of water from collectors in the protected area of Lake Zobnatica, in the Vojvodina region, provided reliable data on the contamination of the selected wetland locality required for effective risk assessments and risk management standards of the selected ecosystem. Additionally, three different water pollution indices were calculated using physicochemical data, offering a comprehensive water quality evaluation in the wetland. Environmental risk evaluation identified five compounds with highrisk quotients: diisobutyl phthalate, dibutyl phthalate, eicosane, phenol, 2,4-bis(1,1-dimethylethyl)-, and diphenyl sulfide. Obtained results would enable a more in-depth understanding of wetland problems such as organic pollution and better management of wetland areas to mitigate pollution problems.

Municipal solid waste management has long been a critical issue for governments worldwide. In recent decades, many countries have seen significant increases in waste generation due to population growth, rapid urbanization, changing lifestyles, and heightened commercial and industrial development. Improper waste management can lead to the leaching of hazardous pollutants, contaminating both ground and surface waters, which can result in environmental harm and adverse health effects for living organisms. This research was conducted through three sampling campaigns in 2022 at a municipal solid waste landfill in Serbia. Gas chromatography-mass spectrometry analysis quantified bisphenol A, diisobutyl phthalate, di(2-ethylhexyl) phthalate, and butylbenzyl phthalate, and projected concentrations in living organisms were calculated using the bioaccumulation factor derived from literature reviews. This study provides a comprehensive overview of the potential effects of these pollutants on both aquatic and terrestrial animals. The aim of this assessment is to initiate discussions on new research directions arising from these findings, highlighting the need for further studies to fully understand the negative impacts of inadequate waste management on human health.

The first case of COVID-19 infection in Serbia was confirmed in early March 2020. State of Emergency was declared soon afterward, imposing shelter in place measures and curfew in the evening hours. A number of countries around the world have experienced air quality improvements during the pandemic, caused by the mandatory quarantine. This study aims to analyze the effect of the quarantine on the air quality, by comparing PM2.5, PM10, and NO2 concentration levels during 32 days quarantine period in the six cities in Serbia, from March 16 to April 16, 2020, with values obtained in the same period in 2019. NO2 concentration levels showed significant reductions in all cities (36-71%). However, PM2.5 and PM10 concentration levels did not show significant reductions in all cities. Concentration levels of PM2.5 and PM10 were lower during the quarantine period only in two cities, by 2 and 19%, and 10 and 30% respectively. Analyzed values can assist decision-makers in defining the levels of effort and scope of actions that need to be undertaken in order to prevent the occurrence of high levels of concentration of PM2.5 and PM10 and to ensure clean and healthy air in Serbia.

Objective: Due to the avid interest in devising economically and environmentally viable and readily available materials for pesticide-contaminated water treatment, the effective-ness of clinoptilolite, a natural zeolite, its modified Na-form, as well as ZSM-5, a synthetic zeolite, was investigated for the removal of the fungicide carbendazim from aqueous solutions. Methods: The removal was investigated using batch ad-sorption by mixing zeolite and the carbendazim solution at different contact times. At the end of each experiment, the suspensions were filtered, and the residual concentration of carbendazim in the filtrates was measured using high-per-formance liquid chromatography (HPLC). The experimental results were used to calculate the efficiency of tested zeolite types for the removal of carbendazim from aqueous solu-tions. Results: Synthetic zeolite was highly effective (94.5%) in removal of carbendazim from aqueous solutions when compared to natural zeolite (19.5–23.3%) and its Na-form (22.5–29.0%). However, the residual carbendazim concen-tration for all three zeolites tested was above the maximum allowable concentration (MAC). Conclusion: Zeolite ZSM-5 exhibited outstanding properties for the removal of the carbendazim fungicide from aque-ous solutions; however, high production costs could be a limiting factor for a wider application in water treatment. For practical applications, natural zeolite may be a more economically viable choice, regardless of its significantly lower removal efficiency. Its surface properties would need to be modified and improved by organic surfactants, which would require additional experiments. Finally, the cost-ef-fectiveness of modified natural zeolite should be compared to the production cost of synthetic zeolite to determine the most effective, economically advantageous, and envi-ronmentally friendly solution for the treatment of pesti-cide-contaminated water.