Increasingly software is taking care of all kinds of decision making for persons and companies alike. At the same time we see the advent of the personal robot. Machines that will take care of mundane tasks like vacuum cleaning or lawn mowing, or that could help our elderly to be less lonely or the sick and disabled to compensate for their inabilities. It is very easy to get exited (or afraid for that matter) of what the future could bring in this regard.
Like the thermostat of your heating system is taking care of the nitty-gritty activities and hence decisions that are needed to control the temperature in your home, robots will start to take care of more and more things that would be controlled and decided upon by humans.
Those robot machines will be helped by the software’s ability to learn and to understand contextual meaning. To help the learning curve to speed up Roboearth has created an open source environment to link up robots. It is based on two pillars: robots can make use of the cloud to get more processing power (i.e. a sort of off-shoring its “brain” while reducing the cost for the robot to be build) but the robots can also share “experiences” as to collaborate in creating “understanding” of the world they are “living” in.
Update 3 March 2014:
via Streamlined rules for robots – MIT News Office.
New algorithms make it easier to write rules for distributed-computing systems, such as networks of sensors, servers or robots.