This study first reviewed previous studies on floor heating systems based on the installation of a phase change material (PCM) and the current status of technical developments and found that PCM-based research is still in its infancy. In particular, the improvement of floor heat storage performance in indoor environments by combining a PCM with existing floor structures has not been subject to previous research. Thus, a PCM-based radiant floor heating system that utilizes hot water as a heat source and can be used in conjunction with the widespread wet construction method can be considered novel. This study found the most suitable PCM melting temperature for the proposed PCM-based radiant floor heating system ranged from approximately 35 °C to 45 °C for a floor thickness of 70 mm and a PCM thickness of 10 mm. Mock-up test results, which aimed to assess the performance of the radiant floor heating system with and without the PCM, revealed that the PCM-based room was able to maintain a temperature that was 0.2 °C higher than that of the room without the PCM. This was due to the rise in temperature caused by the PCM’s heat storage capacity and the emission of waste heat that was otherwise lost to the underside of the hot water pipe when the PCM was not present.

Improvements in the indoor environment at schools are required due to the long periods of time spent in school facilities by children, who are particularly vulnerable to poor indoor air quality (IAQ). This study thus developed an AI‐based model predictive control algorithm (MPCA) for an outdoor air‐based energy recovery ventilation (ERV) system and evaluated its performance. The MPCA was based on artificial neural networks that predicted the particulate matter (PM) concentration, carbon dioxide (CO ₂ ) levels, and temperature to achieve real‐time IAQ control. This algorithm was employed in an elementary school facility, and its performance was empirically compared with a rule‐based algorithm and occupant discretionary control (ODC). All three methods were able to improve the IAQ, with the proposed MPCA outperforming the rule‐based algorithm by maintaining a 21% lower PM _2.5 concentration and 7% lower CO ₂ concentration. In addition, ODC required 2.4–2.5 times higher energy consumption than the other control methods, whereas the MPCA achieved a 5.7% reduction in energy consumption compared with the rule‐based algorithm. Therefore, the MPCA developed in this study is capable of predictive IAQ control for ERV systems and requires less energy than conventional methods, demonstrating its potential to effectively improve the IAQ of school facilities.