Recent advances in robotics have allowed robots to operate in cluttered and complex spaces. Robots, such as humanoids and mobile manipulators, now have the capacity to understand and change the structure of their environments. These robots require new, elaborate planning strategies that handle the full complexity of open-ended, real-world tasks. From long lasting space missions to daily house hold chores, new decision-making methods must handle challenges where semantics and geometry are strongly linked.
Since the development of Shakey, the various fields of planning have made significant progress in their individual domains. Yet, future robot planners must go beyond a simple juxtaposition of symbolic and geometric reasoning. They require new strategies that explicitly consider the interactions between distinct methods of planning. Even within each method, novel data structures, representations and search algorithms should be designed to take into account the complete task: from high-level goals to low level motion commands.
The goal of this workshop is to encourage interaction between researchers in the fields of motion planning, spatial reasoning, task and action planning. We will identify approaches, models and algorithms that bridge the gap between these distinct approaches to autonomous reasoning.
