An academic paper published by a team from the Carnegie Mellon University Human-Computer Interaction Institute has outlined the idea of something the researchers call Zensors. The idea behind Zensors is to use an Android phone and some fancy programming to make the dumb items in your home smart.
Zensors are internet-connected cameras that are able to monitor their surroundings and analyze what they see to send the user data and alerts on pre-set questions that the user wants answers to. An example given is that a camera monitoring a restaurant can be programmed to monitor how many empty drinks are at the bar and send an alert if that number gets too high.
Anything the camera can physically see it could report on. To set the system up the user would circle an area that it wants to be monitored and ask a question in natural language such as “how many empty drinks are there?”
The Zensor cameras can be old Android smartphones or newer internet connected cameras making the sensors easy to deploy. To begin with the system plans to use questions and analysis that is handled by crowd-sourced workers, but the machines are able to learn as the questions are answered by humans. When the machines reach a level of understanding that is good enough, the humans can step out of the picture.