Using algorithms based on finite state machines a method has been developed to identify objects by fixing markers on them. The major advantages of the method are that it works under domestic lighting conditions, in an unstructured surrounding and it is relatively independent of the marker position and orientation. This method is too slow to use in a closed-loop control system to guide a robot, but is ideal to identify static objects and to estimate their position and orientation. This system is intended to form an important sensory information source to a robot at a work station for physically handicapped users that is being developed at Cambridge.