A problem with using expert systems to control robots is the Qualification Problem – trying to predict all of the things in the environment that can prevent the robot from completing the task and developing actions the robot can execute in order to complete the task. Consider the Roomba – the robotic vacuum cleaner. What are three external events that can cause the robot to stop its task and what could we do to allow the robot to solve this problem?
Roomba is a smaller, modernized vacuum cleaner that consequently directs itself around
your home. Like a traditional filter, it grabs earth with turning brushes and a vacuum. There a
side-mounted, thrashing brush that pushes surface underneath the machine and, once there, two
all the more counter-pivoting brushes get the soil and direct it toward the powerful vacuum,
which sucks it away into a little stockpiling container. Not at all like a typical cleaner, has
Roomba moved around your life with two expansive tractor-style wheels, everyone freely
determined by a different electric engine. The wheels can hand over inverse headings, which
implies Roomba can indeed " turn on a dime", and clean any space it can crash into. Control
originates from a locally available NiMH rechargeable battery pack.