Introduction

The patent is a virtual sensor organized with a trained neural network that accepts various parameters measured by sensors that are installed in a smart home for other uses as input.

Technical features

Measurement of environmental CO2 is a difficult and expensive measure. With this patent, a virtual CO2 sensor is proposed which, using temperature, humidity, counting of people in an environment and monitoring the opening of doors and windows, through a neural network trained for this purpose, calculates the CO2 value environmental with good approximation. This measurement is very useful to allow home automation systems to control and adjust environmental parameters by managing the opening or closing of doors and windows or by controlling air treatment systems.  This patented system uses the cooperation of sensors installed in a smart home to extract information content deriving from the contribution of individual sensors. This approach is new and very speculative: in fact, all the applications implemented use various sensors to interpret and have a vision of the observed situation.

Possible Applications

• Control and regulation of ambient air quality;
• Increase in the quality of workplaces and domestic environments;
• Possibility to move motorized windows;
• Allows control of air treatment systems.

Advantages

• Low cost because it uses sensors already installed for other purposes;
• Does not require periodic calibration;
• Use of wireless sensors;
• Placement with any home automation system.