Assembly robots that use an active camera system for visual feedback can achieve greater flexibility, including the ability to operate in an uncertain and changing environment. Incorporating active vision into a robot control loop involves some inherent difficulties, including calibration, and the need for redefining the servoing goal as the camera configuration changes. In this paper, we propose a novel self-organizing neural network that learns a calibration-free spatial representation of 3D point targets in a manner that is invariant to changing camera configurations. This representation is used to develop a new framework for robot control with active vision. The salient feature of this framework is that it decouples active camera control from robot control. The feasibility of this approach is established with the help of computer simulations and experiments with the University of Illinois Active Vision System (UIAVS).

This paper focuses on overcoming the problem of tracking control in structurally flexible redundant manipulators by utilizing their self-motion capabilities. In the proposed algorithm, the self-motion is evaluated in order to nullify the dominant modal force of flexural motion that is induced by a rigid body motion.

The flexure motions of manipulators, which are induced by joint motion, cause undesired inaccuracy in end-effector tracking. In-plath planning states, joint trajectories are so designed as not to excite but to damp out the flexure motions. The self-motion, inherent in redundant manipulators, can alter joint motion, influencing the flexure motion (by exciting and damping the flexure modes), while not affecting end-effector motion at all. Therefore, the self-motion can be utilized to regulate flexibility and effectively reduce the end-effector tracking error.

The effectiveness and applicability of the proposed algorithm have been demonstrated through numerical simulation with three-link planar robotic manipulators possessing flexible links.

Current techniques which have been devised for motion planning and control of compliant multi-fingered grippers are based on the assumption of a constant compliance. The present investigation was undertaken to examine the validity of this assumption. To this end, the variational inequalities approach and the finite element method were used to formulate the general frictional contact problem. A two-step algorithm, which employs Quadratic Programming and non-differential optimization algorithms, is used to solve the resulting variational inequality and to provide the contact characteristics of compliant grippers.

The proposed method is applied to a case study involving a multi-fingered compliant gripper. The equivalent stress contours and the compliance of the system were obtained and analyzed. The proposed approach, which can be applied to any contact problem in robotics, overcomes most of the difficulties and approximations evident in current simplified contact analysis techniques.

In this paper a method for the reduction of the uncertainties in robotised assembly process is presented. The generated assembly strategies are based on the determination of the contact points. The assembly task is divided in two main subtasks, the initial and the main. To reduce the uncertainty of the process a deliberate misalignment of the peg is used.

This paper is focused on the determination of the position and the forces at the contact points between the peg and hole under uncertainty. The available data are the force signal retrieved from the force sensor, the geometry of the assembly parts and the geometry of the robot. The analysis presented in this paper is a part of a knowledge-based system, which generates assembly strategies under the presence of uncertainties.

This paper presents an investigation into the utilisation of sequential and parallel processing techniques for the real-time simulation of a flexible manipulator system. A finite dimensional simulation of the system is developed using a finite difference approximation to the governing dynamic equation of the manipulator. The developed algorithm is implemented on a number of uni-processor and multi-processor, homogeneous and heterogeneous, parallel architectures. A comparison of the results of these implementations is made and discussed, on the basis of real-time processing requirements in the simulation and control of flexible manipulator systems.

The current designs of commercial artificial hands have a low level of innovation. As feedback to the user is difficult to achieve reliably, most devices are simple in design and operation, and limited in functional range. If information on the state of the hand, the forces and any slippage that is occurring is fed back to a microcontroller then more than one degree of freedom can be controlled and a greater and more natural functional range is possible. This paper describes the development of such a device. It outlines the design requirements, the methods of detection of the signals and the training required to operate the hand.

By exploiting an important property of the time optimal control law structure, this paper introduces a new dynamic programming approach for X-Y robot minimum time path following problem. Compared with the conventional dynamic programming method, the proposed approach significantly reduces the computational cost. The efficiency and effectiveness of the proposed approach are demonstrated via computer simulations. In addition, via experimental studies, this paper also addresses practical difficulties in implementing the minimum-time control law.

Reliable flexible assembly workstations depend ultimately on the physics of assembly tasks performed. Multiple-peg insertions, as a practical assembly class, is complicated by the large number and kind of interaction between mating parts.

In this article, geometric and quasi-static analyses of multiple-peg insertions are presented. First, a possible contact-state enumeration, geometric conditions for each contact-state, and the force-moment equations for static-equilibrium states of two dimensional dual-peg insertions are derived. The jamming diagrams and the taxonomy of dual-peg insertions are obtained. Then, an experiment to verify the analysis is presented.

We consider an optimization formulation that captures the quality of a grasp. In this model the object geometry, finger contact points and the magnitude of the external load are given. Perturbation of these parameters from their nominal values might lead to significant variations on the predicted grasping quality measures. In this paper we study the sensitivity of a class of grasping quality functions subject to these perturbations and introduce a global sensitivity measure together with a computational procedure to evaluate it. Our analysis is developed within the framework of sensitivity theory and dual methods of nonlinear programming. Numerical simulations supporting the theoretical analysis are presented.

Manipulators with some flexible links are attractive because they avoid the severe control problems associated with the large inertia forces generated when the large-mass, rigid links in conventional robot manipulators move at high speed. In fact only two of the links within a typical six degrees of freedom revolute-geometry industrial robot cause significant inertia forces, and so only these two links need to be flexible. The development of a two-flexible-link system controller is therefore very relevant to larger manipulators, because it can be readily expanded by adding simple controllers for the other rigid links. Two alternative controllers are developed in this paper, a computed-torque controller and a quadratic optimal controller. Simulations confirm the superior performance of the latter.

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 standard pseudoinverse-based approaches to treat redundant manipulators, the vector of joint increments that corresponds to a desired motion in the space of the secondary task is projected in the Jacobian null space associated with the primary task. In general, this projection may distort the projected vector, so that the secondary task may not adequately be executed. A usual remedy is to rotate the null space projection operator by using a special-purpose weighting matrix. The problem, however, is that this rotation cannot be enforced arbitrarily since it influences the manipulator's performance. In our work we propose an algorithm that is independent on the chosen null space operator and always provides the best attainable motion in the space of the secondary task. Hence, the secondary task is executed more efficiently and the numerical procedure is more robust. A series of numerical experiments confirmed these results.

This paper presents a prototype development of a robot arm to assist walking of elderly people. For effective use in Japanese houses, a robot arm mounted on a rail was designed. Force sensors were used to provide an easy-to-use interface. Thresholds of forces were determined to detect the user's intention to walk as well as for safety considerations. Applying a step by step motion, the arm could follow the rhythmical walking of the subject. We examined an interface method using force sensors and concluded that it can detect the user's motion at the beginning of a walk.

An analytical procedure is proposed for determining the workspace boundary of general telescopic manipulators. The algorithm has been developed by using the Theory of Envelopes providing an algebraic formulation. An analysis of the workspace of telescopic chain has been carried out to emphasize on a general topology and to investigate the effect of link parameters and telescopic stroke.

This paper reports the results of an experimental study, which was conducted to evaluate the performance and implementation aspects of a generalized predictive control (GPC) technique to an electro-hydraulic positioning actuator. Poor dynamics and high nonlinearities form part of the difficulty in the control of hydraulic functions which make the application of adaptive controls an attractive solution. The applicability of GPC to the position control of hydraulic manipulator has been investigated through computer simulations in the literature. However, there is no experimental record of applying this technique to an actual hydraulic system. A suitable plant model is established and recursive U-D factorization technique is adopted for on-line estimation of time-varying plant parameters. Experimental results are obtained from a laboratory electrohydraulic actuator test stand. Various benchmark tests, comprising step and tracking inputs, demonstrate good performance and the promise of the technique. In spite of significant actuator dynamics (control voltage saturation, flow nonlinearity and dry frictional nonlinearity in the hydraulic actuator), successful control tests could be performed repetitively.

This paper reports a visual tracking system that can track moving objects in real-time with a modest workstation equipped with a pan-tilt device. The algorithm essentially has three parts: (1) feature detection, (2) tracking and (3) control of the robot head. Corners are viewpoint invariant, hence being utilised as the beacon for tracking. Tracking is performed in two stages of Kalman filtering and affine transformation. A technique of reducing greatly the computational time for the correlaton is also described. The Kalman filter predicts intelligently the fovea window and reduced computation dramatically. The affine transformation deals with the unexpected events when there is partial occlusion.

The MARK III FAA (Flexible Automatic Assembly) cell was designed to minimise the risk factors associated with FAA. This FAA cell is stepwise upgradeable and adopts the sub-batch principle and reversed material flow solutions. It allows for the assembly of products with an annual volume normally too low to justify automation. The MARK III FAA cell was implemented with the FACE programming and control system, a PC-based, modular programming and control platform for FAA cells. It enables the management of unpredictable order schemes and leads to very low programming costs.

This paper presents a compliant control method for insertion of complex objects with concavities. Most work on robot assembly using compliant motion control schemes focuses on overcoming jamming conditions for simple peg-in-hole problems, and cannot be used for complex shapes frequently encountered in assembly applications. When an object is being inserted to a hole or slot with a small clearance, a contact path is issued to compensate for uncertainties. When the object shape is complex, however, the contact state changes several times and severely, making compliant control difficult. The algorithm presented here is capable of generating satisfactory compliant motion control in spite of changing contact states. During the execution of a nominal motion plan, it computes the actual position of the contact point from the force/torque sensor reading using a contact localization algorithm. It then dynamically updates the center for compliance to the computed contact point, and minimize the chance of jamming and unwanted collisions. The control scheme has been implemented on hardware and tested on the task of inserting a T-shape into a C-shape involving a very tight tolerance. The insertion motion was accomplished by a sequence of 2 translational and 1 rotational compliant motions, and successfully executed by the proposed compliant motion controller.