Force-guided motions of a 6-d.o.f. industrial robot with a joint space approach

This article deals with the interaction between humans and industrial robots, more specifically with the new design and implementation of an algorithm for force-guided motions of a 6-d.o.f. robot. It may be used to comfortably teach positions without using any teaching pendant or for some assistance tasks. For this purpose, from readings of the force/torque sensor mounted in the robot wrist, the gravity forces and torques first have to be eliminated. To control the robot in joint space, it is then convenient to transform the external force and torque values from Cartesian space into joint space using the manipulator transposed Jacobian. This is why with the present approach the Jacobian matrix of the robot used was calculated. Now, from the computed joint torques, suitable position commands of the robot arm can be generated to obtain the desired behavior. A suggestion for this desired behavior is also included in this article. It is based on the impedance control approach in joint space. The proposed algorithm was implemented with the standard Stäubli RX90B industrial robot.     

The effect of handle friction and inward or outward torque on maximum axial push force

OBJECTIVE: To investigate the relationship among friction, applied torque, and axial push force on cylindrical handles. BACKGROUND: We have earlier demonstrated that participants can exert greater contact force and torque in an "inward" movement of the hand about the long axis of a gripped cylinder (wrist flexion/forearm supination) than they can in an "outward" hand movement. METHOD: Twelve healthy participants exerted anteriorly directed maximum push forces along the long axis of aluminum and rubber handles while applying deliberate inward or outward torques, no torque (straight), and an unspecified (preferred) torque. RESULTS: Axial push force was 12% greater for the rubber handle than for the aluminum handle. Participants exerted mean torques of 1.1, 0.3, 2.5, and -2.0 Nm and axial push forces of 94, 85, 75, and 65 N for the preferred, straight, inward, and outward trials, respectively. Left to decide for themselves, participants tended to apply inward torques, which were associated with increased axial push forces. CONCLUSION: Axial push force was limited by hand-handle coupling--not the whole body's push strength. Participants appeared to intuitively know that the application of an inward torque would improve their maximum axial push force. Axial push forces were least when a deliberate torque was requested, probably because high levels of torque exertions interfered with the push. APPLICATION: A low-friction handle decreases maximum axial push force. It should be anticipated that people will apply inward torque during maximum axial push.     

Neural-network based force planning for multifingered grasp

The real-time control of multifingered grasp involves the problem of the force distribution which is usually underdetermined. It is known that the results of the force distribution are used to provide force or torque setpoints to the actuators, so they must be obtained in real-time. The objective of this paper is to develop a fast and efficient force planning method to obtain the desired joint torques which will allow multifingered hands to firmly grasp an object with arbitrary shape. In this paper, the force distribution problem in a multifingered hand is treated as a nonlinear mapping from the object size to joint torques. We represent the nonlinear mapping using artificial neural networks (ANNs), which allow us to deal with the complicated force planning strategy. A nonlinear programming method, which optimizes the contact forces of fingertips under the friction constraints, unisense force constraints and joint torque constraints, is presented to train the ANNs. The ANNs used for this research are based on the functional link (FL) network and the popular back-propagation (BP) network. It is found that the FL-network converges more quickly to the smaller error by comparing the training process of the two networks. The results obtained by simulation show that the FL-network is able to learn the complex nonlinear mapping to an acceptable level of accuracy and can be used as a real-time grasp planner.     

Fast Computation of Optimal Contact Forces

We consider the problem of computing the smallest contact forces, with point-contact friction model, that can hold an object in equilibrium against a known external applied force and torque. It is known that the force optimization problem (FOP) can be formulated as a semidefinite programming problem (SDP) or a second-order cone problem (SOCP), and thus, can be solved using several standard algorithms for these problem classes. In this paper, we describe a custom interior-point algorithm for solving the FOP that exploits the specific structure of the problem, and is much faster than these standard methods. Our method has a complexity that is linear in the number of contact forces, whereas methods based on generic SDP or SOCP algorithms have complexity that is cubic in the number of forces. Our method is also much faster for smaller problems. We derive a compact dual problem for the FOP, which allows us to rapidly compute lower bounds on the minimum contact force and certify the infeasibility of a FOP. We use this dual problem to terminate our optimization method with a guaranteed accuracy. Finally, we consider the problem of solving a family of FOPs that are related. This occurs, for example, in determining whether force closure occurs, in analyzing the worst case contact force required over a set of external forces and torques, and in the problem of choosing contact points on an object so as to minimize the required contact force. Using dual bounds, and a warm-start version of our FOP method, we show how such families of FOPs can be solved very efficiently.     

Sensor Fusion for Compliant Robot Motion Control

Force feedback is necessary for accurate force control in robotic manipulators, and thus far, wrist force/torque (F/T) sensors have been used. But an important problem arises when only these types of sensors are used. In a dynamic situation where the manipulator moves in either free or constrained space, the interaction forces and moments at the contact point and also the noncontact ones are measured by the mentioned sensor. In this paper, an estimator based on a sensor fusion strategy integrating the measurements of three different sensors (a wrist F/T sensor, an inertial sensor, and joint sensors) was developed to determine the contact force and torque exerted by the manipulator to its environment. The resulting observer helps to overcome some difficulties of uncertain world models and unknown environments since it reduces the high-frequency and low-frequency spectral contents, i.e., the low-frequency component due to inertia of a heavy tool mass and the high-frequency component due to impacts. The new improvement was experimentally validated in a force/position impedance control loop applied to a Staubli RX60 industrial robotic platform.     

Muscle responses to simulated torque reactions of hand-held power tools

The aim of this work was to investigate physiological responses to torque reaction forces produced by hand-held power tools used to tighten threaded fasteners. Such tools are used repetitively by workers in many industries and are often associated with upper limb musculoskeletal complaints. The tools considered for stimulation in this study had straight handles and required from 100 to 400 ms to tighten fasteners to a peak torque of 1.0 to 2.5 Nm and from 50 to 150 ms for the torque to decay to zero. A tool stimulator was constructed to apply a programmed torque profile to a handle similar to that of a straight in-line power screwdriver. Wrist flexor and extensor surface EMGs and handle position were recorded as subjects held handles subjected to controlled torque loads that tended to flex the wrist. It was found that: (1) very high EMG values occurred even though torques were of short duration (50 to 600 ms) and the peak torques were low (7–28% of maximum strength); (2) high EMGs in anticipation of torque are directly related to torque build-up rate and peak torque; (3) high peak flexor and extensor EMGs during and following torque onset are related to torque build-up rate and peak torque; (4) minimum time of peak EMGs of 72–87 ms following the onset of torques with 50 ms build-up suggests the contribution of an extensor muscle stretch reflex component; delayed peak for longer build-ups suggests a central control of muscle force in response to torque; (5) angular excursions of handles increase with decreasing torque build-up time and increasing torque magnitude causes increasing eccentric work; (6) the results show that the slow torque build-up times (450 ms) correspond to minimum peak EMGs; and (7) accumulated EMGs increase with increasing torque and torque build-up times. Further studies are needed to evaluate fatigue and musculoskeletal injuries associated with prolonged periods of tool use.     

Hybrid force and motion control of flexible joint parallel manipulators using inverse dynamics approach

An inverse dynamics control algorithm is developed for hybrid motion and contact force trajectory tracking control of flexible joint parallel manipulators. First, an open-tree structure is considered by the disconnection of adequate number of unactuated joints. The loop closure constraint equations are then included. Elimination of the joint reaction forces and the other intermediate variables yield a fourth-order relation between the actuator torques and the end-effector position and contact force variables, showing that the control torques do not have an instantaneous effect on the end-effector contact forces and accelerations because of the flexibility. The proposed control law provides simultaneous and asymptotically stable control of the end-effector contact forces and the motion along the constraint surfaces by utilizing the feedback of positions and velocities of the actuated joints and rotors. A two degree of freedom planar parallel manipulator is considered as an example to illustrate the effectiveness of the method.     

Force capability differences due to gloves

There are many tasks in the industrial environment, involving the use of the hands, for which gloves are required or voluntarily used by many workers. When an individual grasps a handle in attempting to exert short-term maximum force his/ her capability is influenced by the use of gloves.

The purpose of this investigation was to determine the influence of a single glove and a double glove on handle forces generated by individuals. Thirty male volunteers were used as subjects. The maximum pull force, maximum push force, maximum wrist flexion torque and maximum wrist extension torque were measured under the condition of a slippery handle. These exertions of the subject's dominant hand were measured while using no glove, one glove and two gloves. Results indicated that the one-glove condition was superior to the conditions of no glove or double gloves for the forces and torques measured     

Design of a slave arm of a surgical robot system to estimate the contact force at the tip of the employed instruments

Acquisition of the contact force at the instrument tip can enable better performance, e.g. transparency of the haptic feedback in the surgical robot systems. It is, however, difficult to measure the contact force directly due to technical limitations in attaching sensors to the tip of the instruments. This paper proposes a method to estimate the forces by installing the sensors away from the instrument tip. The proposed method employs specially designed mechanical parts of the slave robot, i.e. a slider cover plate for the z-axis translational force along the insertion direction, and docking clamps for the rotational pivot torques around the fulcrum point. Strain gauges are attached to specially designed places with enhancing shapes. The simulation results of the force estimation are presented to confirm the strain concentration area. The proposed method is validated with quantitative experimental results. Calibrated weights are determined upon the comparison of the strain value with a calibrated 6-axis force/torque sensor. The percentage error in the force calibration is about 5~8% calculated by the root mean square error (RMSE) of force-sensing performance. In addition, it can be computed by considering only the bending phase of each sensor although the hysteresis is observed from the calibration graph.     

Initial experiments on a leg mechanism with a flexible geared joint and footpad

In the present work, a leg mechanism together with a flexible joint system is designed and constructed. The leg mechanism is expected to walk on uneven terrain and be subject to certain impact or contact forces. The effect of the designed flexible gear system and footpad on shock absorption is a main concern. The joint torque of the motor gear and the contact force of the foot are analyzed with different control schemes. The control performance of the leg movement by these methods is also presented. The proposed control methodology could next be applied to a newly designed two-link leg mechanism. It is hoped that the present study can be extended to an exoskeleton leg mechanism that is being developed.     

Experiments in contact control

This article describes the implementation, experimentation, and application of contact control schemes for a 7-DOF Robotics Research arm. The contact forces and torques are measured in the sensor frame by the 6-axis force/torque sensor mounted at the wrist, are compensated for gravity, and then are transformed to the tool frame in which the contact task is defined and executed. The contact control schemes are implemented on the existing robot Cartesian position control system at 400Hz, do not require force rate information, and are extremely simple and computationally fast. Three types of contact control schemes are presented: compliance control, force control, and dual-mode control. In the compliance control scheme, the contact force is fed back through a lag-plus-feedforward compliance controller so that the end-effector behaves like a spring with adjustable stiffness; thus the contact force can be controlled by the reference position command. In the force control scheme, a force setpoint is used as the command input and a proportional-plus-integral force controller is employed to ensure that the contact force tracks the force setpoint accurately. In the dual-mode control scheme, the end-effector approaches and impacts the reaction surface in compliance mode, and the control scheme is then switched automatically to force mode after the initial contact has been established. Experimental results are presented to demonstrate contact with hard and soft surfaces under the three proposed control schemes. The article is concluded with the application of the proposed schemes to perform a contact-based eddy-current inspection task. In this task, the robot first approaches the inspection surface in compliance control until it feels that it has touched the surface, and then automatically levels the end-effector on the surface. The robot control system then transitions to force control and applies the desired force on the surface while executing a scanning motion. At the completion of the inspection task, the robot first relaxes the applied force and then retracts from the surface. © 1996 John Wiley & Sons, Inc.     

Guaranteeing prescribed performance and contact maintenance via an approximation free robot force/position controller

In this paper, we consider the problem of force/position tracking for a robot with revolute joints in compliant contact with a kinematically known planar surface. A novel controller is designed capable of guaranteeing, for an a priori known nonsingular initial robot condition, (i) certain predefined minimum speed of response, maximum steady state error as well as overshoot concerning the force/position tracking errors, (ii) contact maintenance and (iii) bounded closed loop signals. No information regarding either the robot dynamic model or the force deformation model is required and no approximation structures are utilized to estimate them. As the tracking performance is a priori guaranteed irrespectively of the control gains selection, the only concern is to adopt those values that lead to reasonable input torques. Finally, a comparative simulation study on a 6-DOF robot illustrates the performance of the proposed controller.     

Observer-Based Dynamic Control of an Underactuated Hand

The purpose of this paper was to construct a velocity observer based on the dynamic model and realize accurate dynamic curve and force control. Curve fitting with the observer obtained precise velocity signals. Compared with PID and factored moment methods, it decreased the fitting errors a lot and achieved ideal results. Compensated with the inverse dynamic equation, the force-based impedance control with the observer could not only realize accurate force tracking, but achieve finger dynamic control by the combination of curve fitting and force tracking. Furthermore, a static grasp model was established for appropriate force distribution. The finger could grasp slippery, fragile, comparatively heavy and large objects like an egg with only base joint torque and position sensors, which illustrated that the hand could accomplish difficult tasks by using the static grasp model and dynamic control.     

Passive torque regulation in an underactuated flapping wing robotic insect

Recent developments in millimeter-scale fabrication processes have led to rapid progress towards creating airborne flapping wing robots based on Dipteran (two winged) insects. Previous work to regulate forces and torques generated by flapping wings has focused on controlling wing trajectory. An alternative approach uses underactuated mechanisms with tuned dynamics to passively regulate these forces and torques. The resulting ??mechanically intelligent?? devices execute wing trajectory corrections to realize desired body forces and torques without the intervention of an active controller. This article describes an insect-scale flapping wing mechanism consisting of a single piezoelectric actuator, an underactuated transmission, and passively rotating wings. Wing stroke velocities are passively modulated to eliminate net airframe roll torque. A theoretical model predicts lift generating wing trajectories and quantifies the passive reduction in roll torque. An experimental structure provides an at-scale demonstration of passive torque regulation.     

Dynamic force/torque balance of 2D polygonal objects by a pair of rolling contacts and sensory‐motor coordination

It is well known that three frictionless fingers suffice to immobilize any 2D object with triangular shape but four fingers are necessary for a parallelepiped. However, it has been recently shown that only two fingers are enough to realize secure grasp of a rigid object with parallel flat surfaces in a dynamic sense if finger ends have a hemispherical shape with appropriate radius and thereby rollings are induced between finger ends and object surfaces. This paper focuses on the two problems: (1) dynamic force/torque balance of 2D polygonal objects under the effect of gravity force by means of a pair of rolling contacts and (2) concurrent realization of dynamically secure grasp and orientation control of 2D polygonal objects by using a pair of multi‐fingered hands with hemispherical ends and sensory feedback signals without knowing object kinematics and mass center. It is shown that the force/torque balance can be attained by controlling both the contact positions and inducing adequate forces in both normal and tangential directions at each of contact points indirectly through finger joints without knowing object mass center and other kinematic parameters. © 2003 Wiley Periodicals, Inc.     

A Study of Reactional Force Compensation Based on Three-Degree-of Freedom Parallel Platforms

Reaction compensation is necessary for space robot applications because reactional forces/torques will cause undesired movement of the spacecraft. Because the reactional torques can be compensated by the existing torque balancing device in the spacecraft, an additional reaction compensating device is necessary to compensate the reactional forces. In this article we study two types of reactional force compensating devices based on three-degree-of freedom, parallel platforms. The first type has three R-P-S legs while the second type has three R-R-P-S leg and a passive R-R-P leg. A three-degree of freedom serial manipulator is used to generate reactional forces, which are to be compensated by the paralle platforms. The kinematics and dynamics of both platforms are analyzed and closed form inverse kinematics solutions are derived. We then design a reactional force compensating device that satisfies the strict volume constraint in a spacecraft. The first type of parallel platform is found to require very long legs due to large orientational motion at certain positions. The second type has smooth motion in both position and orientation, and therefore its size can be very compact. It is concluded that the second type of parallel platform has great potential to be used as a compact reactional force compensating device. © 1995 John Wiley & Sons, Inc.     

Animating human locomotion with inverse dynamics

Because the major force components (the internal muscular forces and torques) are not known a priori over time, you cannot use forward dynamics to predict how the human body will walk. The alternative to the apparently intractable problem of specifying the joint torque patterns in advance is to use inverse dynamics to analyze the torques and forces required for the given motion. Such an analysis can show, for example, that the motion induces excessive torque, that the system is out of balance at a certain point, or that the step length is too great. We present a method of using an inverse dynamics computation to dynamically balance the resulting walking motion and to maintain the joint torques within a moderate range imposed by human strength limits. This method corrects or predicts a motion as indicated by the inverse dynamics analysis. Dynamic correctness is a sufficient condition for realistic motion of nonliving objects. In animating a self-actuated system, however, visual realism is another important, separate criterion for determining the success of a technique. Dynamic correctness is not a sufficient condition for this visual realism. An animation of dynamically balanced walking that is also comfortable in the sense of avoiding strength violations can still look quite different from normal human walking. A visually realistic and dynamically sound animation of human locomotion is obtained using an effective combination of kinematic and dynamic techniques     

Effect of elliptic handle shape on grasping strategies, grip force distribution, and twisting ability

A generic torque model for various handle shapes has been developed and evaluated using experimental data. Twelve subjects performed maximum isometric torques using circular and elliptic cylinders in medium and large sizes (circular: r = 25.4, 38.1 mm; elliptic: semi-major/minor axes = 30.9/19.3, 47.1/27.8 mm) finished with aluminium and rubber, in two opposite directions. Torque, grip force distribution, and finger position were recorded. Maximum torques were 25%, 7%, and 31% greater for the elliptic, large-size, and rubber-finished cylinders than for the circular, medium-size, and aluminium-finished cylinders, respectively. Greater torque for the elliptic cylinders was associated with 58% greater normal force that the subjects could generate for the elliptic than circular cylinders. The model suggests that greater torques for the large-size and rubber cylinders are related to long moment arms and greater frictional coupling at the hand-cylinder interface, respectively. Subjects positioned their hands differently depending on torque direction to maximise their normal force and torque generation. STATEMENT OF RELEVANCE: Desirable handle features for torque generation may be different from those for grip only. Design of handles per advantageous handle features (e.g., shape, size, and surface) may help increase people's torque strength and contribute to increased physical capacity of people.     

Artificial neural network dexterous robotics hand optimal control methodology: grasping and manipulation forces optimization

Optimal fingertip forces can always be computed through the well-known optimization algorithms. However, computation time has always remained a real-time constraint. This article presents an efficient scheme to compute optimal grasping and manipulation forces for dexterous robotics hands. This is expressed as a quadratic optimization problem, and an artificial neural network (ANN) is used to learn such quadratic optimization formulations. Computation has been based on a nonlinear model of fingertip contacts and slips. In achieving object grasping while in motion, the hand Jacobian is considered an important matrix to be computed, but it is also highly intensive for real-time computed applications. Consequently, we investigated an efficient approach using artificial neural networks to learn optimal grasping forces. An ANN is used here to learn the optimal contact forces relating hand joint-space torques to the resulting object force. The results have indicated that the ANN has reduced computation times to reasonable values owing to its ability to map nonlinear force relations. Furthermore, the results have revealed that ANNs are capable of learning highly nonlinear relations relating to distributed fingertip forces and joint torques. The technique developed has also proved to be suitable for off-line learning of computed fingertip forces, even with large training samples.     

Reaction forces, tool handle displacements and discomfort ratings during work with shut-off nutrunners

Reaction forces, hand-arm displacement and subjective ratings were studied during securing of threaded fasteners with nine right-angle nutrunners. The nutrunners had three different preset spindle torques and three different shut-off mechanisms. The tools were tested according to the international ISO method and with a method proposed by the authors in a previous study. Twelve male subjects participated in the study. A force platform measured the reaction forces between the subjects and the floor. Motions of the hand-arm system and shoulder were measured with an optoelectronic measuring system and a goniometer. Strong correlations between motions/ ground reaction forces and ratings were found. The ratings were also strongly correlated with the tool torque impulse, calculated according to ISO 6544, as well as with the time for which the torque exceeded 75% of the preset spindle torque multiplied by the preset spindle torque to produce a 'time-torque value'. The strongest correlation, though, was found with the time-torque value, but the difference was not large enough to propose a change of the ISO standard.