Samsung Pilots Artificial Intelligence for Aging-in-Place

Samsung recently announced it will be piloting an in-home aged care platform that is intended to help elderly Australians age in place at home longer—using technology similar to that found in a self-driving car.

The six-week-long pilot, called The Holly Project, will involve a number of participants who are between 73 and 81 years old and living in five different residential homes in the City of Greater Geelong, Victoria, Australia. Their habits and lifestyle will be tracked utilizing an in-home care platform comprising various software and small hardware-based components, according to ZDNet. The pilot is being done in partnership with Australia’s Deakin University and the City of Greater Geelong.

More specifically, the trial will involve the Holly Smart Home platform, developed by SP Tech Solutions, Samsung’s SmartThings platform, and the Holly Hub, a small computer that receives data from battery-powered sensors located in each participating home. The sensors have been developed to detect motion, temperature, vibration and humidity.

Additionally, speakers will be placed in every room to permit Holly to broadcast audio messages as a means of communicating with participants.

So, how does the technology work?

The Holly Smart Home platform is like a “souped up” smoke alarm, Ian Aitken, Samsung’s director of engineering and solutions, told ZDNet. For instance, Holly will do nothing if there is nothing unusual going on in the home, like a smoke alarm would—but it will communicate to occupants through voice if it detects something unusual. Once this communication occurs, and before Holly sends a notification to designated emergency contacts, the home’s occupants will have three chances to respond through gestures, like waving their arms in the air or opening and closing a door.

“The whole process of Holly is to get to a point that we know what the person is doing in that home and what a normal environment is,” Aitken said. “Once you know what normal is you can flip and know what abnormal is and take action when abnormal takes place.”

At the end of the six-week trial, Holly will be switched off to determine if the participants miss the technology or not. At this time, a wellbeing report for each participant will also be made available to nurses.

Holly’s artificial intelligent (AI) technology is similar to that found in a self-driving car, Rajesh Vasa, a member of Deakin University’s faculty, told ZDNet.

“Holly’s AI has similar learning patterns: There, we’re driving cars and avoiding obstacles, here we’re trying to infer patterns and behavior,” Vasa said. “The advantage is we know it’s a house, people are living in it, and we know people are doing certain things in the house.”

Written by Mary Kate Nelson