Service Robotics in Challenging Environments
For the last six decades, we have witnessed the gradual robotic colonization of industrial spaces with great success. These spaces are often specially adapted for robots, so that they can perform their tasks as precisely and efficiently as possible. Robotics has slowly started its journey from industrial environments to service environments by developing systems to assist humans, typically by performing jobs that are dirty, dull, distant, dangerous, or repetitive.
Service robotics encompass a broad field of applications, most of which having unique designs. Differently to industrial robots, service applications are commonly developed in rich/complex dynamic scenarios, where the robot must explicitly consider uncertain in both perception and action. To deal with such a variety of environments and applications, service robotics span equally grounded, aerial, and underwater systems.
Thus, many of these service applications requires of GPS-denied localization, motion planning in dynamic or partially unknown environments, limited robot perception, specific computer vision skills, limited communications, or complex decision making, to name a few. Although robot autonomy has advance significantly in the last years, many of the solutions to these requirements in the state of the art lack the robustness and reliability required work in challenging environments. Elements as low visibility, fog, dynamic environments, or partial environment information use to degrade the quality of the solutions.
This special session will focus on the robot technologies (perception, planning, localization, actuation) required to develop autonomous or semi-autonomous tasks in challenging environments out of factories, including but limited to:
• GPS-denied localization
• Control in complex environments (slippery terrains, wind turbulences, ocean currents, etc)
• Planning in dynamic scenarios
• Reliable inter-robot communications
• Perception in low visibility
• Mapping or SLAM in low visibility
• Reliable/robust multi-sensor estimations
Robot perception in low visibility environments
Reliable robot localization and SLAM in GPS-denied environment
Robot planning in dynamic and partially known environments
Robot control in complex scenarios
Fernando Caballero, Universidad de Sevila, Spain
Pascual Campoy, Universidad Politécnica de Madrid, Spain
Luis Merino Cabañas, Universidad Pablo de Olavide, Spain
Juan Andrade Cetto, CSIC: Instituto de Robótica e Informática Industrial, Spain
Alberto Ortiz Rodriguez, Universidad de las Islas Baleares, Spain
Arturo de la Escalera Hueso, Universidad Carlos III de Madrid, Spain
José María Armingol Moreno, Universidad Carlos III de Madrid, Spain
Danilo Tardioli, Centro Universitario de la Defensa, Spain