Builders have been talking about making homes smarter for decades. Today, that’s accomplished with systems that monitor energy usage so homeowner’s can reduce the cost of utility bills. What about a house that can think and respond to an owner’s behavioral patterns and adjust the indoor environment accordingly with the use of computers and sensors?
Over the last year an engineering professor at Washington State University, Diane Cook has been studying the limit of what she calls “ambient intelligence” (AI). She and her team have installed sensors in a retirement home in Seattle that detect motion and track the energy usage of resident’s apartments. Computers determine when to raise or lower heat by interpreting these observations.
This new technology where a house has the ability to assess information and act with automating the control of appliances and devices is a far stretch from what smart houses can do today. However, Cook and her team have developed a relatively strong model for such a smart house. Cook envisions the day when homeowners will purchase sensor kits they can install themselves at their local hardware store. The cost to install the wireless sensors and software in a 2,000 square-foot apartment was $2,000.
Cook’s background is in machine learning and data mining. She first gained interest in home automation at the Texas State Fair where she toured the “home of the future.” At the time Cook remembers thinking that what was missing from the idea was intelligence that could transmit information on an occupant’s behavior and refine the home’s operations.
Even though data mining is harder when a room has more than one person in it, Cook doesn’t believe residents will need to wear sensors for optimal readings. She believes the best first step is to use energy monitoring systems so homeowners can get comfortable with the technology. Cook advised the Ars Technicha technology website that homeowners could use their smart phones as a data-gathering tool for the AI system.
Cook’s not just stopping with the philosophy that AI can interpret occupant’s behavior and automate the temperature in the home. She envisions an AI system that will be able to detect cognitive or physical human deterioration based on any changes in occupant behavior. She believes that if the house could provide warnings to occupants, the person who is afflicted may be able to live a longer and healthier life.
Her research has been funded in part by Bosch and Cisco.