The recent success of the Mars Pathfinder mission demonstrates that semi-auto-nomous robotic systems are within reach. A robust and scalable solution to the localisation problem will be central to achieving the goal of complete autonomy. The Pathfinder mission is only one example of a wide variety of application domains for this work. Clearly, the development of autonomous robots will be a significant factor in many domains of exploration, wherever working conditions may present an environment which is hostile to life, or out of the reach of human travel. Examples of such environments include outer space, the depths of the oceans, geothermal hotspots, radioactive or contaminated sites or other extreme environments. Autonomous robots are already in use in automated delivery systems in some hospitals and warehouses and someday we might expect that an autonomous robot will be an integral part of every household: mowing the lawn, vacuuming, or simply tidying up. One significant aspect of realising this goal is that of constructing solutions which are at once practical, efficient and cost-effective. We believe that the work presented here is a step in this direction.