Many recent evidences raised awareness for the Particulate Matter (PM) air pollution with fine (PM2.5), submicrometric (PM1) and ultrafine fractions (PM0.1) and its impact in urban areas, pointing out the adverse consequences on the wellbeing of children, particularly on the respiratory health. The health effects of PM can range from nausea and difficulty in breathing, to skin irritation and include birth defects, significant developmental delays in children, and reduced activity of the immune system, leading to a number of acute or chronic diseases. In this context, the rationale of the research aimed the protection of the children's health related to poor air quality in two cities of Romania i.e. Ploiesti and Targoviste by deploying a PM2.5 monitoring system as a component of an intelligent cyberinfrastructure, which is able to provide early warnings and forecasting of PM pollution episodes. A PM2.5 monitoring microstation prototype containing a measurement, a climate-conditioning, and a control/communication modules, was designed and developed in the ROkidAIR project. The measurement subassembly samples from the atmosphere a relevant air volume, with a controlled flow, filters the 2.5μm fraction, reduces the humidity of intake air volume, passes the airflow through the PM2.5 optical transducer and measures the PM2.5 concentration level at predefined time intervals. Based on the prototype, an advanced PM2.5 monitoring network system comprising 8 microstations provide continuous PM2.5 data in view to be pre-processed and adapted as inputs in forecasting algorithms based on artificial intelligence techniques. The resulted cyberinfrastructure provides information to the end-user terminals quantifying and estimating children's exposure to airborne particulate matter for elaborating early warnings when significant pollution episodes occur.