A.I. eyes smart buildings to predict system failure

Jan 31, 2017, 3:48 AM EST
(Source: Roland Tanglao/flickr)
(Source: Roland Tanglao/flickr)

A Milan-based software firm, CGnal, recently carried out an experiment whereby it used artificial intelligence to help buildings predict failure in critical systems. The company claims that such machine learning algorithm can be of vital importance in safety applications, as it allows identifying faults and fixing them before a facility crashes.

In the experiment, the company analyzed data from heating and ventilation units of an Italian hospital, writes New Scientist. The information was used to train a machine learning algorithm, which picked up the differences in readings of similar appliances. When the software was tested on data from a different duration, it predicted 76 out of 124 real faults.

This is not an isolated effort aimed at harnessing the potential of machine learning, writes BDC Network. Augury, a start-up based in the United States, carries out predictive monitoring by recording audible changes with the help of acoustic sensors installed in machines.