Air pollution is one of the most pressing environmental challenges to public health worldwide. Investigating this problem requires recognizing the distribution of pollution composition and intensity. Therefore, air monitoring systems become more common in big cities worldwide. However, smaller urban centers are frequently overlooked in the literature despite facing similar health risks. The main purpose of this work is to bridge this gap by providing a cost-effective, replicable framework applicable to any lowland town with a limited budget. The proposed monitoring system is a scalable, hybrid network of sensors integrated into the internet of things. The article describes the components of this system and the mathematical method of sensor distribution. The latter is the coarse-graining procedure implemented by weighted k-means clustering based on household emission data. Finally, the application of the proposed system is described by the example of air monitoring in Żary — a lowland town in western Poland. The main results demonstrate the model's high effectiveness: the clustering algorithm successfully identified 15 representative measurement zones, and the subsequent deployment confirmed that combining theoretical spatial optimization with IoT infrastructure yields an accurate, highly representative, and economically viable real-time air quality monitoring system.