Navigating around a house can be difficult for a bot. How does it tell the difference between a wall and an object it should be picking up? In a few years image detection software will change this but right now you will be limited to using fairly basic sensors like ultrasonic, infrared, sound or touch sensors.

People remember where they have been before and where they are going. Moving from one room to another is simple for humans but can be quite challenging for a robot. How does it know if it's in the bathroom or the laundry? We know the difference between a washing machine and the shower. We can also tell the difference between these two things even if we're in someone else's house with a different model washing machine or a different shaped shower enclosure. How can we teach bots to know stuff like that?