Simple Mapping and Path-planning with the Roomba

Reactive architectures have gain wide acceptance as a robust and effective architecture design that yields results through the flexibility of its simplicity. Such architectures have made their way into many commercial products, including the popular iRobot Roomba vacuum cleaner. The existing reactive architecture does not always produce results efficiently and by adding mapping functions and planning to the existing hardware cleaning can be improved without adding expensive hardware that would drive up the cost of the cheap, efficient Roombas. Data points are gathered during normal cleaning runs and later used to probabilistically estimate a map of the cleaning space. The mapping space is then tested by directing the Roomba to move to a remote position in the cleaning space.