Machine learning could help buildings notify occupants about critical systems failures before they happen

Artificial intelligence doesn’t always get the best wrap. From fictional takes on the technology such as Skynet, from The Terminator series, destroying the world and Westworld’s new foreboding take on AI, to Stephen Hawking’s very real warning that artificial intelligence could end mankind, AI certainly doesn’t get as much love as other novel forms of technology like augmented reality or drones.

That doesn’t mean, however, AI isn’t currently being used for very useful applications. For example, a software firm based in Milan, Italy has been attempting to use machine learning algorithms to help buildings predict when their critical systems are about to fail.

According to newscientist.com, CGnal recently analyzed a year’s worth of data, such as temperature, humidity, and electrical use, from heating and ventilation units in an Italian hospital. The data related to appliances in operating theaters, first aid rooms, and corridors.

The firm then trained a machine learning algorithm on data from the first half of 2015 and looked for differences in the readings of similar appliances. When it was tested on data from the second half of the year, the system predicted 76 out of 124 real faults, including 41 out of 44 where an appliance’s temperature rose above tolerable levels, with a false positive rate of 5%, newscientist.com reports.

Augury, a start-up based in the United States, is also experimenting with machine learning. The company uses acoustic sensors that they install in machines to listen for audible changes in function in order to achieve the same result of predictive monitoring. So far, Augury has used the technology to diagnose machines in facilities such as hospitals, power plants, data centers, and a university campus.

By predicting these issues before they occur, facilities could be fixed before a crash ever occurs, saving time and money.

Rural hospitals present unique opportunities and challenges for healthcare facility operators. Oftentimes, the infrastructure and building systems have not been updated for years and require significant improvements in order to meet today’s modern medical demands. Additionally, as these smaller, more remote hospitals are acquired by larger regional and national healthcare systems, the first step by new ownership is often to update and rehabilitate the building. But how can this be done thoughtfully, economically, and efficiently in ways that allow the engineering and facility staff to adapt to the changes? And how can the updates accurately reflect the specific needs of rural communities and the afflictions with which these areas most commonly face?