Special Issue on the Ninth International Workshop on Algorithmic Foundations of Robotics (WAFR)

Robot algorithms are critical building blocks of any robotic system. They enable robots to interpret sensory data, plan courses of actions, control the execution of those actions, and learn new skills, in order to achieve greater autonomy. Today, the design and analysis of robot algorithms are more crucial than ever. • Robotics is undergoing major transformations. Originally focused on industrial manufacturing, it is now rapidly expanding into new domains, such as medical surgery, care of the elderly, ocean and space exploration, transportation, entertainment , homes and offices, and search-and-rescue. In these new domains tasks are less repetitive, environments less structured, and events less predictable. It is impossible to consider explicitly all contingencies in the robot programs. Algorithms that adapt to uncertainties and changes are increasingly needed to achieve greater autonomy over long periods of time. • Robot algorithms are finding applications beyond robotics, for example, in designing mechanical assemblies, model-ing molecular motion, animating digital characters for video games and computer-generated movies, simulating architectural designs, and studying product ergonomy. These non-traditional applications of robot algorithms pose new challenges: dealing with hundreds or thousands of degrees of freedom, handling large populations of individually modeled characters, reasoning under complex physical constraints, and generating natural-looking motions. Solving these challenges will in turn benefit future robotics. Robot algorithms are also rapidly evolving as a result of new technologies , for example, low-cost parallel computers, specialized processors (e.g. GPUs), cheaper and more diverse sensors, new interaction devices ranging from haptic to neuroprosthetic devices, new materials, and miniaturized devices (e.g. MEMS). Unlike most traditional computer algorithms, robot algorithms interact directly with the physical world. They must operate safely, reliably, and efficiently under tight time constraints in imperfectly known environments. So, it is not surprising that the design and analysis of robot algorithms raises unique combinations of fundamental questions in computer science, electrical engineering, mechanical engineering, and mathematics. For example, minimalist robotics studies the minimal sensing and actuation capabilities required to complete a given task. It addresses not only computational complexity issues, but also " physical " complexity issues. Probabilistic methods are widely used as a modeling tool to handle uncertainties due to sensing and actuation noise, but they are also increasingly used as a computational tool to avoid prohibitively expensive computations by sampling partial information and handling the uncertainties that result from the lack of a " complete " computation. The Workshop on Algorithmic Foundations of Robotics …