Smart Home Knows Just How You Like Your Breakfast
From: New Scientist - 09/02/2009
By: MacGregor Campbell

Washington State University researchers have created a sensor-filled smart
apartment that can learn the routines of its inhabitants by observing how
they walk around the home and use appliances. The researchers say the
technology could be used in homes to help people with cognitive difficulties
or dementia perform their daily routines. For example, the apartment can
recognize when a person is making breakfast, and if the person accidentally
leaves the stove burner on, the system can detect the anomaly and provide
audio and video signals to tell the person to turn the burner off. The
computer system that analyzes the sensors' output, known as Casas, can learn
a person's habits without prior assumptions on what events or patterns to
expect. The researchers have successfully tested the system in a
single-resident apartment on campus. The system needed about a month of
observation before it could identify the resident's habits. Casas uses
sensors that detect motion, temperature, light, humidity, water use, door
contact, and the use of items such as appliances. Data-mining algorithms also
were created to help the sensors, which are less accurate than cameras and
other detection systems. One algorithm uses a grid of motion sensors to map
out how a person walks around the home, finding daily routes through the
house, while another algorithm finds patterns in a sequence of events, such
as learning to expect the resident to turn on a tap after turning on the
oven. A third algorithm correlates events it detects with the time of day to
identify patterns, such as when a person eats. 

Read the entire article at:
http://www.newscientist.com/article/dn17724-smart-home-knows-just-how-you-like-your-breakfast.html

Links:
Casas
http://ailab.wsu.edu/casas/

Parisa Rashidi
http://eecs.wsu.edu/~prashidi/index.html

Diane Cook
http://www.eecs.wsu.edu/~cook/