Humans exhibit outstanding learning, planning and adaptation capabilities\nwhile performing different types of industrial tasks. Given some knowledge\nabout the task requirements, humans are able to plan their limbs motion in\nanticipation of the execution of specific skills. For example, when an operator\nneeds to drill a hole on a surface, the posture of her limbs varies to\nguarantee a stable configuration that is compatible with the drilling task\nspecifications, e.g. exerting a force orthogonal to the surface. Therefore, we\nare interested in analyzing the human arms motion patterns in industrial\nactivities. To do so, we build our analysis on the so-called manipulability\nellipsoid, which captures a posture-dependent ability to perform motion and\nexert forces along different task directions. Through thorough analysis of the\nhuman movement manipulability, we found that the ellipsoid shape is task\ndependent and often provides more information about the human motion than\nclassical manipulability indices. Moreover, we show how manipulability patterns\ncan be transferred to robots by learning a probabilistic model and employing a\nmanipulability tracking controller that acts on the task planning and execution\naccording to predefined control hierarchies.\n