Obtaining the optimal cost-to-go map for large scale rough terrains is computationally very expensive both in terms of duration and memory resources. A fast algorithm for approximation of the optimal cost-to-go map in terms of terrain traversability measures for path planning on known large scale rough terrains is developed. The results show that the majority of the cost-to-go map values, computed from every terrain location with respect to the goal location, are near-optimal. Unlike Dijkstra algorithm, the proposed algorithm has inherently parallel structure, and can be significantly speeded up depending on the number of used CPU cores.

This paper presents a novel navigation and motion planning algorithm for mobile vehicles in rough terrains. The main purpose of the algorithm is to generate feasible trajectories while selecting smoother paths, in the sense of level of roughness, toward the goal position. The purpose is achieved by adapting the passivity-based model predictive control optimization setup (PB/MPC), recently proposed for flat terrains, to the case of an outdoor irregular terrain. The passivity-based concept is used to enhance MPC in order to stabilize the goal position guaranteeing the task completion. The framework which is obtained can exploit any vehicle model in order to carefully take into account the vehicle dynamics and terrain structure as well as the wheel-terrain interaction. The inherited property of the MPC optimization allows to impose any additional constraint into the PB/MPC navigation, such as those needed to prevent vehicle rollover and unnecessary sideslip. The cost function representing the level of roughness along a candidate path is used to select the appropriate terrain areas toward the goal position. The results have been verified by several simulation examples.

Torque sensors in robotics are used in those applications where high precision positioning of the end effector is required. Technological difficulties often arise in the mechanical design as well as in the placement of the sensor. However, the use itself of the torque sensor for feedback control, in all those cases where elasticity of the transmission is an issue, is maybe not completely understood. The main goal of this paper is to put the control problem in a clear and correct setting, both from the steady state and from the dynamic standpoints. The role of the zero in the origin of the complex plane in the transfer function from motor torque to output torque is thoroughly discussed. It is also shown, through analysis and simulations on a detailed model of the space robotic manipulator DEXARM, that a correct design of the torque loop allows to overcome some performance limitations in the control of the load position, that arise in the case of simple positional control.

In this paper, an integral friction model is proposed, defining separately the Dahl, the Stribeck and the microviscous effects. The model is based on an integral closed form solution of the Dahl model. It is shown by simulation that the model is consistent with the LuGre model, while being computationally more efficient. A limit solution of the proposed model avoids the nonphysical drifts, a drawback of the LuGre model recently addressed in the literature.

The stability properties of an operational space controller applied to a space robotic manipulator are investigated in this paper. It is shown that a major potential source of instability is the drive train dynamics in the joints of the manipulator. An analysis carried out with all normalized parameters shows that a nominally stable position control system may turn out to be unstable in practice. Problems arise also for force control, if proportional controllers are used. The analysis is supported by simulations obtained on a detailed dynamic model of the SPIDER arm, to be used in a space program on the International Space Station.

An impedance controller for industrial manipulators is presented in this paper. Special attention has been paid to all the aspects that qualify an industrial robot, including decentralized proportional-integral-derivative position control, torsional flexibility, and friction at the joints. A discussion of a single-degree-of-freedom (DOF) case is first presented. The selection of the impedance parameters is based on the behavior in contact and makes reference to a new design parameter, the compliance bandwidth. The single-DOF analysis is then used in the design of the impedance controller for a complete manipulator. A rotational impedance based on a geometrically consistent representation of the orientation is used. Several experimental results obtained on a 6-DOF industrial robot, equipped with a force/torque sensor, are presented. The experiments show the effectiveness of the proposed approach in several conditions, including contact with an extremely stiff surface and noncontact/contact transition.

A new impedance controller is presented. Special attention has been paid to all those aspects that qualify an industrial robot, including decentralized PID positional control, torsional flexibility and friction at the joints. Discussion of a single degree of freedom case proved to be valuable in the design of the impedance controller for a complete manipulator. Several experimental results obtained on a 6 degrees-of-freedom industrial robot equipped with a force/torque sensor, are given.

Describes work on a large laboratory testbed for space robotics (Columbus Automation and Robotics Laboratory Testbed), able to execute hierarchically organized complex activities, to increase the degree of autonomy of the system. For instance, now the robotic system is able to conclude several activities, like closing a drawer or moving a sample from a location to another, without any intervention of the operator even when the subject to be manipulated is out of its locking place or is far from the expected (nominal) position of several centimetres. A set of laser distance sensors available in the end-effector has been exploited to this purpose, and suitable strategies have been developed to recover the true grasping positions. The work has been developed following a standard control development methodology which has led to the full modular integration of the additional autonomous control capabilities in the basic robot control and programming system.

Permanent magnet AC motors generate parasitic torque pulsations, owing to several electromagnetic phenomena. Excitation of the mechanical resonances on the load side is generally the worst consequence of these disturbances. Speed oscillations arise which may dramatically limit the performances in high precision applications. In the paper a control scheme which autonomously identifies the parameters of the disturbance model, through simple closed loop motion experiments, is outlined. Experiments carried out on an industrial robot demonstrate the effectiveness of the control scheme in suppressing oscillations, both on the motor and on the load side.

This paper describes the results of an experimental research on the disturbances acting on the measured force during contact motion of industrial robots. Two phenomena have been found to be mainly responsible for these disturbances, namely the arm structural elasticity due to the torsional flexibility of the joints and the torque ripple of the brushless motors. The structural elasticity is characterized by a resonance frequency function of the arm configuration, while the torque ripple has a principal harmonic of frequency proportional to the motor velocity. Though obtained on a particular experimental setup, these results should be of general interest, since both joint flexibility and torque ripple are common drawbacks of most industrial robots.

With the goal of filling the gap between theory and industrial applications, an implicit hybrid control scheme is proposed in this paper, designed to fit as much as possible the conventional industrial robot control architecture. The dynamic effects due to joint compliance, which are a major source of performance degradation in industrial robots, are fully taken into account. The scheme is based on a task description particularly suited for a direct integration in conventional robot programming tools and aims at exerting the force control action without affecting the trajectory tracking. Only the differential kinematic model (Jacobian) of the robot is needed in the design of the force control law, while the force control loop is charged with rejecting dynamic disturbances due to motion. A thorough experimental validation of the strategy, both in terms of force regulation and trajectory tracking capabilities, is discussed, based on experiments performed on an industrial robot, endowed with a six-axis wrist force/torque sensor and with a laser distance sensor.

This paper presents the Automation and Technology testbed for External Servicing (ARTES) developed by Tecnospazio on behalf of European Space Agency Technology Center (ESTEC). ARTES was conceived as a ground testbed to perform technology assessment of

The dynamic mathematical model of a redundant industrial manipulator is described. The model has been obtained by exploiting the features of a continuous-system modular simulation code. These features make it possible to bypass the difficulties related to closed-mechanical chain modeling. A partial validation of the model has been carried out on the basis of experimental test results. Experimental and simulation data are shown and compared.>