PARALLEL SESSION - Architecture

Enhancing air quality in intelligent buildings

Jun 4, 2024 | 3:05 - 3:25 PM
SAN GIUSTO ROOM

Indoor air quality in home and offices has a strong impact on human health and productivity. There are raising concerns of human exposure to indoor concentrations of chemicals, with most studies on real building environments considering carbon dioxide as the most important pollutant. However, nowadays, attention goes also to other pollutants as radon, particulate matter (like PM2.5) and other volatile organic components like the protections used for fabrics or the varnish from wooden furniture.
Many different trade-offs are usually considered during the design stage of the buildings in order to properly address indoor environmental quality, considering parameters like air ventilation and comfort for example. However, it is not always easy to predict the dynamic changes that can take place during the operation phase of the buildings, which can include for example modifications in the expected pattern of occupancy of the offices, changes in the destination of common areas, or even large modifications in the office layouts.
With the wide availability of IoT sensors, it is now possible to quantify the air pollutants concentrations in buildings. It is clear that the larger the number of sensors deployed, the better the accuracy of the physical parameters. However, the installation, deployment and maintenance of the network of sensors can be not only very costly, but also very time consuming.
This presentation describes a sensor placement data-driven approach which in a very user-friendly approach, provides a means to graphically analyse the complex interactions between the different types of sensors. While the use of the proposed technique provides no guarantee of Pareto optimal placement, it provides an easy strategy to manage the day-to-day operational changes in building occupancy. The approach is based on a technique called Self Organizing Maps (SOM), a kind of artificial neural network that is trained with an unsupervised learning algorithm. While the name and the definition of the technique can sound intimidating for a non-expert, the user-friendly graphical interface provided by modeFRONTIER makes the technique extremely easy to use.

Subscribe to our newsletter

Stay up to date with news, events, upcoming webinars and innovative applications.

This field is required.
This field must be a valid email address.

By clicking I accept the privacy policy

Thank you for your subscription!
There was a problem with your subscription

Seems there was a problem with your subscription, please check the form fields or try again later.