The smart home will alert you to emergencies Bashny.Net

CGnal firm, located in Milan, recently analyzed the annual volume of data from heating and ventilation in the Italian hospital. Now the sensors permanently embedded in the air conditioning, heating and ventilation, engineers analyzed data on temperature, humidity and use of electricity in operating theatres, wards ambulance and corridors.





They trained a machine learning algorithm on data from the first half of 2015, so that he was looking for differences in the readings from the various instruments. Then tested the algorithm on data from the second half of the year will the program to predict failures before they happen? The system predicted 76 of 124 real-world crashes, including 41 of the 44 when the temperature rose above a set level.



"We started with hospitals because the heating, ventilation and air conditioning system there is critical, says Carlo Annis, who worked on the experiment. These predictive algorithms will help to eliminate problems before they happen. However, there is a definite problem, as the differences in functionality become less noticeable with aging of the devices, and the difference between normal and failure is becoming more and more indistinguishable.

Such systems predictive maintenance are now engaged not only in Italy but also in USA and in Finland.published

 

Source: ecotechnology