Special issue on Ground robots operating in dynamic, unstructured and large-scale outdoor environments: Editorial

International audienceReal-­world outdoor applications of ground robots have, to date, been limited primarily to remote inspection of suspected explosive devices and, with less success, to the broader domain of remote survey and inspection in hazardous environments. Such robots have almost exclusively been tele-­operated. Also notable as examples of outdoor ground robots are the planetary rovers, currently deployed with great success on the surface of Mars. But with the rapid development of autonomous (driverless) cars, and the emergence of robotic vehicles in agriculture, it is likely that there will be significant growth in both the numbers and scope of commercial ground robots in outdoor environments in the near future. Applications in real-­world outdoor environments demand many different capabilities of robotic systems: from autonomy, decision-­‐making and sensing to robustness, reliability and the ability to manage limitations on communications and power. Outdoor robotic contests such as the European Land Robot Trial (ELROB) or the euRathlon competition present challenges in which participants compete in real world scenarios under realistic conditions and in real terrains, which have not been specially prepared beforehand and are unknown to the competing teams. Example scenarios include autonomous navigation over several kilometres, environmental survey or search and rescue in terrains with dense forests, hilly landscapes with scattered and varied vegetation and water ponds – day and night and with unpredictable weather conditions. For this special issue we called for papers that present land robot systems deployed in the field in similar realistic challenges. We sought papers that focus on any aspect of robotic systems, from vehicle design to the overall system architecture and control, via terrain mapping, localization, mission planning and execution – with an emphasis on systems that fulfil a specific real world task. We specified that robot or system innovations must be supported by extensive field results. Also that field tests must be under realistic and challenging conditions with respect to the terrain type, the scenario to be achieved, and/or the conditions within which the scenarios must be achieved. A total of eleven papers were submitted to this special issue, and following a rigorous process of anonymous peer-­review six have been selected for publication. The papers included in this special issue cover a broad spectrum of the challenges outlined above. Two papers are dedicated to 3D mapping and localization, two present an autonomous navigation system, and two present wholly integrated systems that operate in populated areas. Outlined below, the six papers of this special issue strongly represent the state-­of-­the-­art in this important area of research