1. Ivan Stevović, Innovation Center of the Faculty of Mechanical Engineering, University of Belgrade, Belgrade, Republ, Serbia
2. Mihailo Jovanović, Faculty of Management Herceg Novi, University Adriatik, Montenegro, Montenegro
3. Jovana Jovanović, Faculty of Civil Engineering and Management, University Union Nikola Tesla, Belgrade, Serbia
Environmental monitoring increasingly depends on dense sensor networks, remote platforms and automated analytics capable of transforming heterogeneous measurements into timely knowledge for decision making. This paper examines how artificial intelligence and sensor theory can be integrated to improve the reliability, interpretability and usefulness of environmental monitoring systems. The central argument is that AI can enhance detection, prediction and response only when it is coupled with rigorous sensor theory, including calibration, uncertainty quantification, sampling design and validation against reference observations. The paper reviews applications in air, water, soil, climate and ecosystem monitoring, emphasizing Internet of Things architectures, machine learning, deep learning and remote sensing. It also discusses methodological requirements for data preprocessing, sensor fusion, model selection and uncertainty-aware inference. The results indicate that AI-enabled monitoring can increase spatial and temporal resolution, reduce operational costs and support proactive environmental management, but these benefits depend on transparent models, robust calibration, representative training data and governance frameworks that protect data quality and public trust. The paper concludes that sustainable environmental monitoring should be designed as an integrated socio-technical system in which sensors provide traceable measurements, AI provides adaptive interpretation, and human expertise ensures accountability and context-sensitive decisions.
Ključne reči :
Tematska oblast:
Datum:
10.08.2026.
Contemporary Materials 2026 - Savremeni Materijali
Datoteka uz rad
(1.25 MB)
