We've already seen assistive robots for seniors, which sit in one place in the user's home. A team at Washington State University, however, has developed one that figures out where its user is, then goes to them and offers its assistance when needed.

The university's Robot Activity Support system (RAS) consists of a mobile wheeled robot that operates within a previously-designed "smart" apartment. By wirelessly accessing a network of sensors within that dwelling, the robot is reportedly able to determine where the resident is, what they're doing, and whether they need help.

If it finds that its assistance is needed, it makes its way through the home in order to get to the person, utilizing its mapping and navigation camera to avoid obstacles as it goes. Once it reaches them, it uses a tablet interface to offer one of three things – it can show them a video illustrating how to perform the next step in the task at hand, it can show a video of the entire task, or it can lead them to objects necessary to complete the task.

In a test of the system, the robot was used to help 26 undergraduate and graduate students complete three tasks within the apartment. Those tasks included taking a dog out for a walk, taking medication with food and water, and watering plants. Whenever the students were having difficulty with any of these, the apartment's sensor system detected the problem, and responded by summoning the robot to their location.

Most of the volunteers found the subsequent assistance experience to be favorable, and thought the tablet interface was easy to use. The scientists are now planning on getting senior volunteers to assess the system.

"Upwards of 90 percent of older adults prefer to age in place as opposed to moving into a nursing home," says lead scientist Prof. Diane Cook. "We want to make it so that instead of bringing in a caregiver or sending these people to a nursing home, we can use technology to help them live independently on their own."

A paper on the research was recently published in the journal Cognitive Systems Research.

View gallery - 2 images