There is increasing interest in using rehabilitation robots to assist post-stroke patients during rehabilitation therapy. The motion control of the robot plays an important role in the process of functional recovery training. Due to the change of the arm impedance of the post-stroke patient in the passive recovery training, the conventional motion control based on a proportional-integral (PI) controller is difficult to produce smooth movement of the robot to track the designed trajectory set by the rehabilitation therapist. In this paper, we model the dynamics of post-stroke patient arm as an impedance model, and propose an adaptive control scheme, which consists of an adaptive PI control algorithm and an adaptive damping control algorithm, to control the rehabilitation robot moving along predefined trajectories stably and smoothly. An equivalent two-port circuit of the rehabilitation robot and human arm is built, and the passivity theory of circuits is used to analyze the stability and smoothness performance of the robot. A slide Least Mean Square with adaptive window (SLMS-AW) method is presented for on-line estimation of the parameters of the arm impedance model, which is used for adjusting the gains of the PI-damping controller. In this paper, the Barrett WAM Arm manipulator is used as the main hardware platform for the functional recovery training of the post-stroke patient. Passive recovery training has been implemented on the WAM Arm, and the experimental results demonstrate the effectiveness and potential of the proposed adaptive control strategies.

Robot-aids or Rehabilitators are our chosen neologism to name a new class of robotic devices that represent a substantially departure from prior applications of robotics in rehabilitation. Rather than use robotics as an assistive technology for a disabled individual, we envision robots and computers as supporting and enhancing the productivity of clinicians in their efforts to facilitate a disabled individual's recovery. In this paper, we attempt a brief overview of our work in what promises to be a ground breaking field. We discuss the concept of robot-aided neuro-rehabilitation as a means to deliver therapy, measure patient performance, and also as a design tool. To illustrate the broad spectrum of neurological diseases that this technology might impact, we will illustrate each case with a different pathology, namely cerebral vascular accident (CVA – also known as stroke), Parkinson's disease (PD), and cerebral palsy (CP).

In this paper the concept of a serial-parallel macro-milli manipulator is presented for medical applications by investigating the feasibility of combining two different robotic structures into one system. The combination of the two structures can be used as a redundant system in which stiffness, singularity and accuracy can be optimized as related to workspace demands and restrictions that are required for a priori selected medical applications. A case study has been experienced at LARM (Laboratory of Robotics and Mechatronics) at the University of Cassino through a new system composed of a commercial Adept COBRA and CaPaMan2bis (Cassino Parallel Manipulator 2bis). The performance of this system is compared with a previous serial-parallel macro-milli robotic system also developed at LARM in Cassino. Experimental tests have been carried out in order to characterize the proposed systems.

The estimation of dynamic parameters in mechanical systems constitutes an issue of crucial importance both for inverse dynamics based control strategies and dynamic simulation applications where high accuracy is required. The identification procedures can be classified in two main groups: indirect and direct procedures. The first ones act sequentially in several steps in each of them parameters of different nature (basically friction and inertial parameters) are identified by means of specifically designed experiments, while the direct procedures allow the identification of all parameters defining de dynamic model in a single stage. In this paper, the implementation and comparison of an indirect and a direct identification procedures on an industrial robot provided with an open control architecture is addressed.

This paper presents an approach for analyzing the dynamics of initial impact during robotic parts mating represented by peg-in-hole assembly. Lagrange's impact model is used to derive a general form of impact equations for an industrial manipulator performing peg-in-hole assembly. Specific impact equations can then be generated for a given robot to analyze part motions and contact forces in the mating of parts. The impact equations of a SCARA robot are obtained and utilized to investigate how the system parameters affect force impulse and departure angle in the assembly of a peg with a chamfered hole. The resulting information is useful to robot manufacturers in design of robotic equipment and to robot users in determining optimal operational conditions. It is also useful for building an intelligent robot controller capable of deciding on proper values of operation parameters.

In this paper, we propose a bioinspired path planning algorithm for finding a high-quality initial solution based on the pipeline of the Rapidly exploring Random Tree (RRT) method by modifying the sampling process. The modification mainly includes controlling the sampling space and using the probabilistic sampling with the two-dimensional Gaussian mixture model. Inspired by the tropism of plants, we use a Gaussian mixture model to imitate the tree’s growth in nature. In a 2D environment, we can get an approximate moving point’s probabilistic distribution, and the initial path can be found much quickly guided by the probabilistic heuristic. At the same time, only a small number of nodes are generated, which can reduce the memory usage. As a meta-algorithm, it can be applicable to other RRT methods and the performance of underlying algorithm is improved dramatically. We also prove that the probabilistic completeness and the asymptotic optimality depend on the original algorithm (other RRTs). We demonstrate the application of our algorithm in different simulated 2D environments. On these scenarios, our algorithm outperforms the RRT and the RRT* methods on finding the initial solution. When embedded into post-processing algorithms like the Informed RRT*, it also promotes the convergence speed and saves the memory usage.

In this paper neural network (NN) control techniques for non-model based PD controlled robot manipulators are proposed. The main difference between the proposed technique and the existing feedback error learning (FEL) technique is that compensation of robot dynamics uncertainties is done outside the control loop by modifying the desired input trajectory. By using different NN training signals, two NN control schemes are developed. One is comparable to that in the FEL technique and another has to deal with the Jacobian of the PD controlled robot dynamic system. Performances of both controllers for various trajectories with different PD controller gains are examined and compared with that of the FEL controller. It is shown that the new control technique performed better and robust to PD controller gain variations.

A fundamental study on the control of hydraulically actuated robots is presented. Dynamic modelling is performed in both time-domain and frequency-domain. It is shown that the inclusion of hydraulic elements increases the order of the system. Hydraulic compliance is the most effective factor in this regard.

Three distinct control strategies are applied. Their performances are evaluated and compared. All three methods are exemplified with a two link hydraulic robot in a computer simulation. The robot has the same hydraulic configuration as many existing industrial manipulators. The simulation program is written in ACSL (Advanced Continuous Simulation Language) running on a VAX 11/750.

The paper considers a technique for computation of the inverse kinematic model of the human arm for robot based rehabilitation that uses measurements of the hand position and orientation and radial acceleration of the upper arm. Analytical analysis and empirical validation of the method are presented. The algorithm enables estimation of human arm angles, which can be used in trajectory planning for rehabilitation robots, evaluation of motion of patients with movement disorders, and generation of virtual reality environments.

The paper is aimed at generating optimal swing motions during the single-support phase of sagittal gait. Unlike the previous Part 1 which deals with passive motions, all joints of the biped are assumed to be active in the present Part 2. The final conditions specify an impactless heel-touch in order to avoid a destabilizing effect on the biped motion. As the biped is essentially submitted to gravity forces, the motion is generated by minimizing the joint actuating torques. Feasible motions are defined by state inequality constraints limiting joint motions, and defining foot clearance and obstacle avoidance during the swing. The optimization problem is dealt with using Pontryagin's Maximum Principle. A final two-point boundary value problem is solved by implementing a shooting method. The approach presented is illustrated by various numerical simulations applying to a seven-body planar biped which has four or five active joints during the swing phase.

The accuracy of pose of industrial robots is often unsatis-factory for advanced applications. Particularly regarding off-line programming, exchangeability and high precision tasks problems may occur which can be very time-consuming and costly to solve. Therefore a calibration system ROSY has been developed in order to increase the accuracy of standard robots and parallel-kinematic structures, like the Tricept robots.

Similar to the traditional adaptive strategies for robot manipulators, the regressor-free adaptive controller design also requires applying Slotine and Li's modification to avoid the feedback of joint accelerations. In this paper, a simple method is proposed to construct a regressor-free adaptive controller for robot manipulators without Slotine and Li's modification. In the new design, the joint acceleration vector is lumped into an unknown time-varying function and the function approximation technique is utilized to cover its effect; therefore, its implementation is free from joint acceleration feedback. The closed-loop stability and boundedness of internal signals are justified by the Lyapunov-like technique. Both simulation and experimental results for a two-link robot are presented to show the effectiveness of the proposed design.

This paper presents studies of the coordination of human upper body voluntary movement. A minimum-jerk 3D model is used to obtain the desired path in Cartesian space, which is widely used in the prediction of human reach movement. Instead of inverse kinematics, a direct optimization approach is used to predict each joint's profile (a spline curve). This optimization problem has four cost function terms: (1) Joint displacement function that evaluates displacement of each joint away from its neutral position; (2) Inconsistency function, which is the joint rate change (first derivative) and predicted overall trend from the initial target point to the final target point; (3) The non-smoothness function of the trajectory, which is the second derivative of the joint trajectory; (4) The non-continuity function, which consists of the amplitudes of joint angle rates at the initial and final target points, in order to emphasize smooth starting and ending conditions. This direct optimization technique can be used for potentially any number of degrees of freedom (DOF) system and it reduces the cost associated with certain inverse kinematics approaches for resolving joint profiles. This paper presents a high redundant upper-body modeling with 15 DOFs. Illustrative examples are presented and an interface is set up to visualize the results.

A novel approach for trajectory tracking of a mobile-robots formation by using linear algebra theory and numerical methods is presented in this paper. The formation controller design is based on the formation states concept and the dynamic model of a unicycle-like nonholonomic mobile robot. The proposed control law designed is decentralized and scalable. Simulations and experimental results confirm the feasibility and the effectiveness of the proposed controller and the advantages of using the dynamic model of the mobile robot. By using this new strategy, the formation of mobile robots is able to change its configuration (shape and size) and follow different trajectories in a precise way, minimizing the tracking and formation errors.

Autonomous landing on the deck of a boat or an unmanned surface vehicle (USV) is the minimum requirement for increasing the autonomy of water monitoring missions. This paper introduces an end-to-end control technique based on deep reinforcement learning for landing an unmanned aerial vehicle on a visual marker located on the deck of a USV. The solution proposed consists of a hierarchy of Deep Q-Networks (DQNs) used as high-level navigation policies that address the two phases of the flight: the marker detection and the descending manoeuvre. Few technical improvements have been proposed to stabilize the learning process, such as the combination of vanilla and double DQNs, and a partitioned buffer replay. Simulated studies proved the robustness of the proposed algorithm against different perturbations acting on the marine vessel. The performances obtained are comparable with a state-of-the-art method based on template matching.

This paper proposes an experimental approach for evaluating the backlash error of an ABB IRB 1600 industrial serial robot under various conditions using a laser interferometer measurement instrument. The effects of the backlash error are assessed by experiments conducted on horizontal and vertical paths. A polynomial model was used to represent the relationship between the backlash error and the robot configuration. A strategy based on statistical tests was developed to choose the degree of polynomial representing the effect of the tool center point (TCP) speed and payload. Results show that the backlash error strongly affects the repeatability of the industrial robot. Statistical analyses prove that the backlash is highly dependent on both robot configuration and TCP speed, whereas it remains nearly unaffected by changes in the payload. It was discovered that the backlash error as measured at the TCP may exceeds 100 μm, and that the positive backlash error increases and the negative backlash error decreases when there is increase in TCP speed.

The robotic intervention has great potential in the rehabilitation of post-stroke patients to regain their lost mobility. In this paper, firstly, we present a design of a novel, 7 degree-of-freedom (DOF) upper limb robotic exoskeleton (u-Rob) that features shoulder scapulohumeral rhythm with a wide range of motions (ROM) compared to other existing exoskeletons. An ergonomic shoulder mechanism with two passive DOF was included in the proposed exoskeleton to provide scapulohumeral motion with corresponding full ROM. Also, the joints of u-Rob have more range of motions compared to its existing counterparts. Secondly, we propose a fractional sliding mode control (FSMC) to control u-Rob. Applying the Lyapunov theory to the proposed control algorithm, we showed the stability of it. To control u-Rob, FSMC has shown effectiveness to handle unmodeled dynamics (e.g. friction, disturbance, etc.) in terms of better tracking and chatter compared to traditional SMC.

This paper discusses the study of a spherical parallel manipulator (SPM) used as a haptic device for tele-operation applications. The SPM presents poor behavior in singular configurations. Redundancy is used to eliminate the parallel singularity without major changes in the mechanical structure. Comparisons in terms of kinematic and dynamic behavior between the non-redundant and redundant SPM are presented. The results prove the advantage of introducing redundancy in the actuator and sensor to improve the behavior of the SPM. A new control strategy for the redundant SPM is also proposed. The control strategy has been successfully tested and validated on a SimMechanics model.

The paper addresses kinematic and geometrical aspects of the Orthoglide, a three-DOF parallel mechanism. This machine consists of three fixed linear joints, which are mounted orthogonally, three identical legs and a mobile platform, which moves in the Cartesian x-y-z space with fixed orientation. New solutions to solve inverse/direct kinematics are proposed, and we perform a detailed workspace and singularity analysis, taking into account specific joint limit constraints.

Researchers dream of developing autonomous humanoid robots which behave/walk like a human being. Biped robots, although complex, have the greatest potential for use in human-centred environments such as the home or office. Studying biped robots is also important for understanding human locomotion and improving control strategies for prosthetic and orthotic limbs. Control systems of humans walking in cluttered environments are complex, however, and may involve multiple local controllers and commands from the cerebellum. Although biped robots have been of interest over the last four decades, no unified stability/balance criterion adopted for stabilization of miscellaneous walking/running modes of biped robots has so far been available. The literature is scattered and it is difficult to construct a unified background for the balance strategies of biped motion. The zero-moment point (ZMP) criterion, however, is a conservative indicator of stabilized motion for a class of biped robots. Therefore, we offer a systematic presentation of multi-level balance controllers for stabilization and balance recovery of ZMP-based humanoid robots.