The construction industry is a major contributor to dust, greenhouse gases, and other air pollutants. Implementing effective and sustainable practices in managing construction site operations can greatly mitigate the environmental effects of a project. To achieve this, a collaboration between a scientific research institution and a construction company enabled the real-time monitoring of air quality parameters at a construction site using Internet of Things (IoT) technologies. They implemented an IoT-based system framework that integrated a distributed sensor network to collect real-time data from the construction site. Various sensors were utilized to gather data on the concentration of NO₂ and particulate matter (PM_2.5 and PM₁₀), as well as meteorological parameters such as wind speed, wind direction, humidity, pressure, and temperature. The real-time measurements yielded insights into the level of air pollution at the construction site and its association with earth excavation, the primary construction activity. This information can be utilized to manage excavation work and reduce the levels of polluting gases (NO₂) and suspended particles. By conducting an on-site monitoring of these three pollutants, the study discovered that the dust levels resulting from excavation activities were relatively high. When comparing the wind direction with NO₂ and PM concentrations, it was concluded that earth excavation significantly influenced the air quality in the construction area. However, in terms of the primary factors affecting NO₂ and construction dust concentrations, the analysis revealed that meteorological factors did not exhibit a significant correlation with NO₂ and dust levels at the construction site. The multiple linear regression (MLR) and the artificial neural network (ANN) models for predicting PM_2.5, PM₁₀ and NO₂ concentration in air using meteorological parameters as predictors were applied. The ANN model showed greater accordance with the measured concentrations in air than the MLR model.

The importance of masonry arch bridges as a traffic network element calls for a thorough analysis focused on both structural stability and loading capacity of these historical structures, considering the usage of these bridges in contemporary traffic conditions. This paper focuses on the analysis of longitudinal cracks in a single span masonry arch bridge to evaluate its influence on structural behaviour of the system. As longitudinal cracks do not necessarily present an inevitable collapse mechanism, analysis of the causes is crucial for evaluating the serviceability and functionality of the bridge investigated. The methodology is based on the following: literature review, observation of the stone bridge in Ivanjica, geological testing of the site, geophysical testing of the bridge, laboratory testing of mechanical characteristics of stone used for the bridge construction and biological analysis of the samples of implemented materials on the bridge. Finite element analysis of the bridge was conducted to define the causes of the longitudinal cracks. The 3D simulation model was based on the data collected through observation and experimental analysis. This paper provides extensive research on a single span masonry bridge, examining how different deterioration mechanisms, in conjunction, can lead to the appearance of cracks in masonry arch bridges and provide remedial measures accordingly.

In current practice, the remediation of landslides has shown that the biggest problem is the increase in the number of works, and therefore the price of the works. This is due to several factors, including characteristic of the soil, such as the collapse (collapse) of the surrounding ground around the main slide during landslide remediation. Unless these soil erosion effects are taken into account, recovery costs will overrun, which can jeopardize the planned budget. This paper presents a multi-criteria optimization of landslide remediation using the PROMETHEE method and determines the optional number of walls for the additional soil erosion. In a case study on examples of real landslides in the Republic of Serbia, the application of the method is presented and appropriate conclusions are drawn.

Accurate prediction of pile base resistance is essential for the safe and economical design of deep foundations, particularly in soft soils where load-transfer mechanisms are highly nonlinear and uncertain. This study develops a comparative, probabilistic, and interpretable machine learning framework for predicting pile unit base resistance using five input variables: applied load, settlement, effective pile length, axial stiffness, and SPT value. A Gaussian Process Regression model with an automatic relevance determination (ARD) Exponential kernel achieved the best performance, with RMSE = 262.11 kPa, R² = 0.943 on an independent test set, and 95% prediction intervals with 96.46% coverage. Beyond record-level evaluation, a leave-one-pile-out validation (the first grouped validation applied to this database) showed harder generalization to entirely unseen piles, driven mainly by a per-pile level offset rather than shape mismatch (within-pile correlation = 0.975). A sequential next-stage scheme, calibrating this level from a pile’s early loading stages, then predicted its remaining segments with consistently strong agreement (Willmott’s d = 0.76–0.83), supporting practical extension of partial load tests. Interpretability was assessed using ARD, SHAP, permutation/ablation importance, and partial dependence/accumulated local effects analysis, identifying settlement as the dominant predictor. The framework combines accuracy, calibrated uncertainty, interpretability, and validated segment-level extrapolation for reliability-oriented pile assessment.