Showing papers by "Paolo Rocco published in 2015"

A pre-collision control strategy for human-robot interaction based on dissipated energy in potential inelastic impacts

Abstract: Enabling human-robot collaboration raises new challenges in safety-oriented robot design and control. Indices that quantitatively describe human injury due to a human-robot collision are needed to propose suitable pre-collision control strategies. This paper presents a novel model-based injury index built on the concept of dissipated kinetic energy in a potential inelastic impact. This quantity represents the fracture energy lost when a human-robot collision occurs, modeling both clamped and unclamped cases. It depends on the robot reflected mass and velocity in the impact direction. The proposed index is expressed in analytical form suitable to be integrated in a constraint-based pre-collision control strategy. The exploited control architecture allows to perform a given robot task while simultaneously bounding our injury assessment and minimizing the reflected mass in the direction of the impact. Experiments have been performed on a lightweight robot ABB FRIDA to validate the proposed injury index as well as the pre-collision control strategy.

Safety-aware trajectory scaling for Human-Robot Collaboration with prediction of human occupancy

Abstract: Planning and control of an industrial manipulator for safe Human-Robot Collaboration (HRC) is a difficult task because of two conflicting requirements: ensuring the worker's safety and completing the task assigned to the robot. This paper proposes a trajectory scaling algorithm for safe HRC that relies on real-time prediction of human occupancy. Knowing the space that the human will occupy within the robot stopping time, the controller can scale the manipulator's velocity allowing safe HRC and avoiding task interruption. Finally, experimental results are presented and discussed.

Reactive Task Adaptation Based on Hierarchical Constraints Classification for Safe Industrial Robots

Abstract: A widespread and flexible use of robots in rapidly changing working environments could be greatly enhanced by human–robot interaction and collaboration. Humans and robots have complementary skills. The robotic worker can relieve the human from repetitive work, and the human can make robot deployment easier by managing nonstandard or particularly skilful operations. Such a scenario, however, requires new safety systems to preserve human workers from potential danger and at the same time to make human–robot interaction productive and advantageous. In this paper, a system for safe and task consistent human–robot interaction integrated with an industrial controller is proposed. The robot executes evasive motions to avoid impacts with obstacles consistently with the task. A classification of constraints constituting the task is proposed and a safety strategy based on such classification is defined. This paper finally presents integration of the safety system with an industrial controller and experimental validation on an assembly operation.

Reactive motion planning and control for compliant and constraint-based task execution

Abstract: In this work, we propose a constraint-based algorithm for combined trajectory generation and kinematic control for robotic manipulators. The main feature of the algorithm is to ease robot programming, shifting from an imperative paradigm, in which task constraints are semantically and uniquely mapped into a suitable end-effector velocity profile, towards a declarative motion programming, where such constraints are turned by the controller into motion commands only at run-time: The system embeds the capability of handling real-time events, such as updated sensor readings, with reduced pre-programmed control logics. An experimental case study based on a 7-DOF robot demonstrates the effectiveness of the approach.

Constraint-based Model Predictive Control for holonomic mobile manipulators

Abstract: In this paper, a controller based on constrained optimization for tracking problems in mobile manipulation is presented. A Model Predictive Control problem is set and solved online, allowing to deal with dynamic scenarios and unforeseen events. Besides acceleration, velocity and position constraints, collision avoidance constraints for the mobile base and the arm and Field-of-View constraints have been enforced and extended over the prediction horizon. Navigation performance has been improved by including an additional goal, derived from the classical vortex field approach, to the MPC problem. An experimental validation on a KUKA youBot mobile manipulator has been carried out, showing the online applicability of the presented approach.

Passivity-based control of robotic manipulators for safe cooperation with humans

Abstract: This paper presents a novel approach to the control of articulated robots in unstructured environments. The proposed control ensures several properties. First, the controller guarantees the achievement of a goal position without getting stuck in local minima. Then, the controller makes the closed-loop system passive, which renders the approach attractive for applications where the robot needs to safely interact with humans. Finally, the control law is explicitly shaped by the safety measure – the danger field. The proposed control law has been implemented and validated in a realistic experimental scenario, demonstrating the effectiveness in driving the robot to a given configuration in a cluttered environment, without any offline planning phase. Furthermore, the passivity of the system enables the robot to easily accommodate external forces on the tool, when a physical contact between the robot and the environment is established.

A redundancy resolution method for an anthropomorphic dual-arm manipulator based on a musculoskeletal criterion

Abstract: In order to make humans feeling comfortable when working with robots, it is necessary for robots to be as much as possible “human-like” in both their appearance and movements. In redundant manipulators, it is possible to use the additional degrees of freedom in order to make the robot motion more human-like, thus increasing the quality of the humanrobot interaction. In this work, a redundancy resolution method to address this issue is presented. Such a method considers the human musculoskeletal system and a biomechanical model of the human upper limbs in order to define a strategy to solve the redundancy for a dual-arm anthropomorphic manipulator as a human would do, then making the robot able to both perform the prescribed task and to assume human-like postures.

Detecting, tracking and predicting human motion inside an industrial robotic cell using a map-based particle filtering strategy

Abstract: In order to enable safe and efficient human-robot interaction it is beneficial for the robot control system to be able not only to detect the presence and track the motion of human workers entering the robotic cell, but also to predict in the least possible time their trajectory and the area they are heading to This paper proposes an innovative particle filtering strategy addressing at the same time the problems of Human Detection and Tracking and Intention Estimation, based on low-cost commercial RGB surveillance cameras, a map of the robotic cell environment, and a probabilistic description of the trajectories followed by human workers inside the cell Results of several validation experiments are presented

Estimating a Mean-Path from a set of 2-D curves

Abstract: To perform many common industrial robotic tasks, e.g. deburring a work-piece, in small and medium size companies where a model of the work-piece may not be available, building a geometrical model of how to perform the task from a data set of human demonstrations is highly demanded. In many cases, however, the human demonstrations may be sub-optimal and noisy solutions to the problem of performing a task. For example, an expert may not completely remove the burrs that result in deburring residuals on the work-piece. Hence, we present an iterative algorithm to estimate a noise-free geometrical model of a work-piece from a given dataset of profiles with deburring residuals. In a case study, we compare the profiles obtained with the proposed method, nonlinear principal component analysis and Gaussian mixture model/Gaussian mixture regression. The comparison illustrates the effectiveness of the proposed method, in terms of accuracy, to compute a noise-free profile model of a task.