Skip to main content
We are incredibly proud to celebrate a remarkable achievement by Mikki with the publication of his research in the Young Scientist Journal.

We are incredibly proud to celebrate a remarkable achievement by Mikki with the publication of his research in the Young Scientist Journal.

Mikki’s study explores how artificial intelligence can be used to improve the performance and efficiency of HVAC (Heating, Ventilation, and Air Conditioning) systems while maintaining precise environmental conditions.

Maintaining stable temperature and humidity is critical in many industries, particularly pharmaceutical manufacturing, where even small variations can impact product quality and regulatory compliance. Traditional machine learning approaches for HVAC control often depend on large amounts of historical data and ideal operating conditions. In practice, however, data may be limited, incomplete, or not fully representative of real-world system behavior.

To address this challenge, Mikki developed an innovative AI architecture that separates the control process into two neural networks working together. The first network predicts optimal actuator settings, such as water control valve positions and air bypass damper adjustments, based on environmental conditions. The second network predicts the resulting room temperature and humidity levels generated by those control actions. This approach allows the system to better understand both the decisions being made and their expected outcomes.

The research was trained using two years of real-world operational data collected at five-minute intervals from a pharmaceutical manufacturing facility in Thailand. By learning from actual industrial conditions rather than simulations alone, the system was able to develop a more practical and robust understanding of HVAC performance.

The results demonstrated that the proposed approach could reduce Characterized Control Valve usage while maintaining an average room temperature of 23.6°C and relative humidity of 59.4%. These findings suggest that AI-driven control systems can help facilities operate more efficiently while continuing to meet strict environmental requirements.

This publication reflects Mikki’s curiosity, dedication, and passion for applying emerging technologies to solve real-world engineering challenges. His work contributes to the growing field of intelligent building automation and highlights the potential of artificial intelligence to improve energy efficiency and sustainability across industries.

We are thrilled to see Mikki’s efforts recognized through this publication and excited to follow his future accomplishments. Congratulations, Mikki, on this outstanding achievement from us all at NatMatSci.

Read Mikki’s paper here and press play below to hear him talk about his research and the journey behind this exciting achievement.