Reproductive medicine has recently witnessed significant advancements, particularly in vitro fertilization (IVF). One crucial aspect of IVF involves the extraction of cellular material and its analysis to maximize the chance of successful implantation. This work highlights the development and application of the automated system for Trophectoderm cell (TE) extraction and separation, addressing the need for precision, efficiency, and reduced manual intervention. The presented automated system is equipped with a computer vision algorithm, microliter pump, vacuum system, and micromanipulation tools to consistently and accurately biopsy TE cells. An experimental setup is developed to verify the behavior of the proposed method, in which a holding micropipette is connected to a vacuum system and holds the embryo stationary. Three steps are performed to complete the process and are controlled by a computer vision algorithm. The coordinates of the Zona Pellucida (ZP) perforation (perforated in a previous step) are used as a feedback signal to a simple proportional controller to control the biopsy pipette motion. The computer vision monitors the amount of TE cells aspirated inside the biopsy pipette and controls the microliter pump. The aspirated TE cells were separated away using a laser cutting system. Experimental results demonstrate that the system can relocate the biopsy pipette, TE cell extraction, and separation.

In this paper, we present a precise robot-aided single-cell surgery system to perform single-cell biopsy for cells <25 μm in diameter. A microfluidic chip is designed to arrange upto 100 individual cells in an array. A micropipette mounted onto a 3-DOF micromanipulator and a computer mouse-operated high-precision XY stage is developed to perform high-precision and high-throughput single-cell biopsy. The system is evaluated experimentally by extracting two organelles from adherent cells patterned in a microfluidic chip. The fluorescent-labeled nucleus and mitochondria of human foreskin fibroblast cells are biopsied to demonstrate the capability of the proposed system. The survival rate of the semi-automated biopsy is 73% and 45% for mitochondrial and nucleus biopsies, respectively.

Micromanipulation of individual biological cells, such as embryos during preimplantation genetic diagnosis (PGD), is a delicate and time-consuming task. Two major procedures in PGD include the rotation of the embryo to gain a more favorable position for zona breaching, and the extraction of a blastomere after an opening has been made in the zona pellucida. Rotation tracking of cleavage-stage embryos has not been reported given their lack of distinctive features. In this manuscript, a geometric model for partially determining the three dimensional (3-D) angular position of 2-cell embryos using two dimensional (2-D) microscopic brightfield images was derived and verified using a computer generated model. This model was then applied using computer vision algorithms on a rotating cleavage-stage mouse embryo, demonstrating partial 3-D rotation tracking. Furthermore, embryo micromanipulation tasks are typically performed manually using micropipettes. Technological advances have made automation of these tasks possible. This manuscript also presents computer vision algorithms for the segmentation and calibration of micropipettes. The calibration procedure allowed automated position control of the micropipettes without the need for real-time vision feedback using micropipette recognition algorithms, and effective position control was verified using semi-automated blastomere extraction experiments. This manuscript presents preliminary work towards the automation of cell manipulation procedures.

Research and biomedical applications in cell surgery require transportation and rotation of biological cells. In these cell manipulation tasks, the cell of interest must be translated and oriented properly such that the desired component, such as the polar body or other organelles, can be imaged with optical microscopy. This paper presents a holographic optical tweezers (HOT) based system to carry out automated translational control in the plane, and rotational control about one rotational axes of a suspended cell. Based on the proposed general equations of motion of the cell, held in an optical trap, two controllers, one for cell translational and one for rotational control, are developed to translate and orient the cells to the desired position and orientation in a sequential manner. Experiments are performed to demonstrate the effectiveness of the proposed approach.

Optical tweezers has become a powerful tool in automated cell transportation control and has been used in a variety of biological applications. The use of optical tweezers for cell surgery has great potential for various biomedical applications such as microinjection, organelle extraction and modification, and preimplantation genetic diagnosis (PGD). In these cell surgical manipulation tasks, the cell of interest must be oriented properly such that the desired component, e.g., the polar-body or organelles, can be visualized by optical microscopy; thus cell rotation becomes a necessary procedure. Currently, cell rotational control can be carried out by laser tools that are usually handled by skilled people. The open-loop manual operation cannot be readily used for applications requiring precise and high throughput cell rotational control. This highlights the need of developing an automated controlled robot manipulator to rotate biological cells more accurately and efficiently. In this paper, we propose a cell surgery system that utilizes two optical traps, generated by robotically controlled holographic optical tweezers (HOT), to manipulate the cell for rotation, where the optical tweezers functions as special robot manipulators. Through dynamic modeling using T-matrix approach, the relationship between the applied torques and the spherical coordinates of the optical tweezers is characterized. A rotational controller is further developed to rotate the cell to the pre-desired orientation accurately. Experiments are performed to demonstrate the effectiveness of the proposed approach.

Automatic manipulation and patterning of biological cells into an artificial scaffold is an imperative step in the production of high-quality tissue for tissue transplantation. This paper examines the incorporation of dielectrophoresis into a three-dimensional (3D) scaffold body for batch manipulation and patterning of cells. To facilitate dielectrophoresis-based manipulation, a multi-layer biocompatible scaffold structure utilizing its body as the integrated micro-electrodes was designed and fabricated using soft lithography. Voltage of opposite polarity was applied to the scaffold structure and the resultant electric field from the scaffold body polarized the cells in the culture medium and attracted them to migrate towards the scaffold body. Experiments were conducted and the results confirm that the proposed multi-layer scaffold is capable of generating dielectrophoretic forces to manipulate the cells to the scaffold surface, forming a three-dimensional cellular pattern automatically.

This paper presents the fabrication of a three-dimensional microcoil through the technique of microassembly. The microcoil design is comprised of nine out-of-plane micro-sized windings. Each winding was assembled onto the base substrate orthogonally by a robotic manipulator through microassembly. In contrast to the conventional serial pick-and-place microassembly, this work incorporated the approach of parallel microassembly to grasp and assemble three windings onto the base substrate simultaneously for increased productivity. In addition, a vision-based algorithm was developed to automate the parallel grasping process of three windings. This algorithm utilized well-defined templates to provide high-precision position and orientation evaluations for the micro-sized components. The performance of the microcoil fabrication process was evaluated and discussed. To establish better electrical contact between the windings and the base substrate, conductive adhesive was introduced in the assembly process and the electrical properties of the assembled microcoil structure were examined.

We present a fully automated micrograsping methodology that uses a micro-robot and a microgripper to automatically grasp a micropart in three-dimensional (3-D) space. To accurately grasp a micropart in 3-D space, we propose a three-stage micrograsping strategy: (i) coarse alignment of a micropart with a microgripper in the image plane of a video camera system; (ii) alignment of the micropart with the microgripper in the direction normal to the image plane; (iii) fine alignment of the micropart with the microgripper in the image plane, until the micropart is completely grasped. Two different vision-based feedback controllers are employed to perform the coarse and fine alignment in the image plane. The vision-based feedback controller used for the fine alignment employs position feedback signals obtained from two special patterns, which can achieve submicron alignment accuracy. Fully automated micrograsping experiments are conducted on a microassembly robot. The experimental results show that the average alignment accuracy achieved during automated grasping is approximately ± 0.07 μm; the time to complete an automated micrograsping task is as short as 7.9 seconds; and the success rate is as high as 94%.

In this paper, a vision-based impedance control algorithm is proposed to regulate the cell injection force, based on dynamic modeling conducted on a laboratory test-bed cell injection system. The injection force is initially calibrated to derive the relationship between the force and the cell deformation utilizing a cell membrane point-load model. To increase the success rate of injection, the injector is positioned out of the focal plane of the camera, used to obtain visual feedback for the injection process. In this out-of-plane injection process, the total cell membrane deformation is estimated, based on the$X-Y$coordinate frame deformation of the cell, as measured with a microscope, and the known angle between the injector and the$X-Y$plane. Further, a relationship between the injection force and the injector displacement of the cell membrane, as observed with the camera, is derived. Based on this visual force estimation scheme, an impedance control algorithm is developed. Experimental results of the proposed injection method are given which validate the approach.

Biological cell injection is laborious work that requires lengthy training and suffers from a low success rate. In this paper, a robotic cell-injection system for automatic injection of batch-suspended cells is proposed. To facilitate the process, these suspended cells are held and fixed to a cell array by a specially designed cell-holding device, and injected one by one through an ldquoout-of-planerdquo cell-injection process. A micropipette equipped with a polyvinylidene fluoride microforce sensor to measure real-time injection force is integrated in the proposed system. Through calibration, an empirical relationship between the cell-injection force and the desired injector pipette trajectory is obtained in advance. Then, after decoupling the out-of-plane cell injection into a position control in theX-Yhorizontal plane and an impedance control in theZ-axis, a position and force control algorithm is developed to control the injection pipette. The depth motion of the injector pipette, which cannot be observed by microscope, is indirectly controlled via the impedance control, and the desired force is determined from the onlineX-Yposition control and cell calibration results. Finally, experimental results demonstrate the effectiveness of the proposed approach.

Automatic cell injection has been the focus of many researches and commercial development for several years. In this paper, a robotic cell injection system for automatic batch injection of suspended cells is developed. To facilitate the process, these suspended cells are held and fixed to a cell array by a specially designed cell holding device, and injected one by one through an “out-of-plane” cell injection process. A micropipette equipped with a PVDF micro force sensor is integrated in the proposed system. The force sensor is utilized to measure real time injection force applied to the cells during injection process. Through calibration of the relationship between the cell injection force and the desired injector pipette trajectory, a position (vision) and force control algorithm is proposed and applied to the motion control of the injection pipette in three-coordinate directions during an injection process. The out-of-plane cell injection task is decoupled into a position control in X-Y horizontal plane and an impedance force control in Z-axis. The depth motion of the injector pipette, a common problem of three-dimensional micromanipulation, is indirectly controlled by the force control. Finally, experimental results are given to demonstrate the effectiveness of the proposed approach.

Biological cell injection is laborious work which requires lengthy training and suffers from a low success rate. Even a tiny excessive manipulation force can destroy the membrane or tissue of the biological cell. This makes the control of the injection force an important factor in the cell injection process. In this paper, a vision-based impedance force control algorithm is proposed based on dynamic modeling of a laboratory test-bed injection system. The injection force is calibrated in a cell injection task to derive the relationship between the force and the cell deformation. A cell biomembrane point-load model is utilized in this force calibration. In three-dimensional cell injection task, the total cell membrane deformation is estimated, based on the X - Y coordinate frame deformation of the cell, as measured with a microscope, and the known angle between the injector and the X Y plane. Further, a relationship between the injection force and the injector visual displacement of the cell membrane is derived. Based on this force visual estimation scheme, an impedance force control algorithm is developed. Finally, experimental results are given which demonstrate the effectiveness of the proposed approach.

In this paper, we propose a vision control strategy to perform automatic microassembly tasks in threedimension (3-D), and develop relevant control software. Specifically, using a 6 degree-of-freedom (DOF) robotic workstation to control a passive microgripper to automatically grasp a designated micropart from the chip, pivot the micropart, and then move the micropart to vertically insert into a designated slot on the chip. In the proposed control strategy, the whole microassembly task is divided into two subtasks, micro-grasping and micro-joining, in sequence. To guarantee the success of microassembly and manipulation accuracy, two different two-stage feedback motion strategies, the pattern matching and auto-focus method are employed, with the use of vision-based control system and the vision control software developed. Experiments conducted demonstrate the efficiency and validity of the proposed control strategy.

In this paper, to study the effect of different control approaches on improving trajectory tracking accuracy for a 3 degree-of-freedom (DOF) planar parallel robot, we tested two synchronized-type controllers: PI-type synchronized control and adaptive synchronized (A-S) control; and conventional PID control and adaptive control. Here PID control and PI-type synchronized control are dynamic model-free while the adaptive control and A-S control are dynamic model-based. Because of the closed-loop kinematic chain mechanism of the experimental planar parallel robot used in this study, trajectory tracking control of this robot may be treated as a synchronization problem, and consequently, use of the synchronized control approaches can substantially improve the trajectory tracking performance of the robot end-effector compared with approaches without synchronization. Through conducting experiments on an experimental 3-DOF P-R-R type planar parallel robot by using the four control approaches, the above claims are demonstrated.

This paper describes a new visual force sensing method for measuring microforces acting upon the jaws of passive microgrippers used in the assembly of MEMS devices. The importance of force measurement during microassembly is to confirm the microgripper-micropart make a successful grasp and to also protect the microparts and microgripper from excessive forces which may lead to damage during the assembly process. In the proposed approach, the force measurement problem is reduced to a problem of determining the microgripper jaw displacement. A 3-D finite element model is developed to study the relation between the forces and the displacements. The resulting nonlinear force-displacement relationship is fitted into a second degree equation. Computer vision is used to measure the relative displacements of the right and left microgripper jaws with respect to the microgripper base during assembly. Patterns that were introduced to the microgripper during the design phase are used to measure those relative displacements through pattern identification. Two-dimensional pattern identification is performed using normalized cross correlation template matching, to estimate the degree to which the image and pattern are correlated

This paper describes the design and development of a 6 degree of freedom robotic manipulator used in the assembly of three-dimensional MEMS (micro electromechanical systems) microstructures. The robot employs a highly innovative mechanical design for the rotational axes to provide unprecedented access to a microchip substrate for microassembly operations. The first three axes of the robotic manipulator are orthogonally mounted linear stages providing Cartesian positioning of the chips beneath the end effector (microgripper). A rotational stage (alpha) mounted on the distal end of these three Cartesian axes allows the MEMS chip to be rotated. Two more degrees of freedom (beta and gamma) are serially mounted to the base frame, allowing for two degrees of rotation of the end effector. This configuration permits assembly of micro-parts on the surface of a MEMS chip at any orientation angle to the surface, within the limits of the workspace of the manipulator and the resolution of the motors. The end effector employs a standard tungsten probe with a passive microgripper bonded to it, which is used for grasping micro-parts. A software system has been developed to allow automatic operation of the manipulator. Preliminary assembly tests confirm the usefulness of the proposed design

In this paper, in order to improve tracking accuracy and satisfy multiple closed-loop performance specifications simultaneously during high-speed, high-acceleration trajectory tracking for a 3 degree-of-freedom (DOF) planar parallel manipulator, we propose a new control approach, termed convex synchronized (C-S) control. The C-S control is base on the so-called convex combination method and synchronized control. To implement the synchronized control scheme, a feedback signal, termed the synchronization error, is employed, which represents the degree of coordination of the active joints in the parallel manipulator based on the kinematics of the parallel manipulator, and thus tracking accuracy is improved. On the other hand, the convex combination method utilizes the convex property of the required closed-loop performance specifications. Through combining multiple linear controllers, so-called sample controllers, each of which satisfies at least one closed-loop performance specification, a C-S controller is algebraically calculated, which satisfies all closed-loop performance specifications simultaneously. Compared with traditional trial-and-error method, the convex combination method is more straightforward and efficient. Hence, possessing advantages of the synchronized control and the convex combination method, the proposed C-S control method can both improve tracking accuracy and satisfy multiple closed-loop performance specifications simultaneously (MSS). Experiments conducted on a 3-DOF P-R-R type planar parallel manipulator demonstrate the above claims

In this paper, a new motion control hardware architecture is proposed for improved motion performance of robotic manipulators during high speed motion. The main idea is to remove the servo control loop from the DSP (digital signal processor) to a FPGA (field programmable gate array), and utilize the high speed hardwired logic of the FPGA to enhance the computation capability and relieve the computing load on the DSP. The control algorithm is partitioned into a linear portion and a nonlinear portion. The linear portion with position/velocity feedback represents the major control loop and is implemented in the FPGA. The nonlinear portion acts as dynamic compensation to the linear portion to perform complex modeling related calculations, and is implemented in the DSP. A new FPGA-based motion control IC is designed to realize this new control hardware structure. Experiments were conducted on a Yamaha robot manipulator to compare new control architecture and the existing one, when the same control algorithm was used. Experimental results demonstrate that the proposed new control architecture exhibits much improved motion performance especially during high-speed motions

A newly designed lightweight planar parallel manipulator is designed to improve operational speed of electronic manufacturing processes and implement a "smart parallel manipulator" through the integration of a parallel mechanism architecture and active control of linkage vibration using lead zirconate titanate (PZT) transducers. Boundary conditions and mode shapes of intermediate linkage are not conventional due to linkages undergoes constrained rigid body motion. Through numerical simulations using a substructuring dynamic model developed for the parallel manipulator, the boundary condition is determined close to a pinned-pinned boundary condition. This conclusion is confirmed using experimental modal analysis (EMA) using a random motion input. However, it is observed that linkage vibration exhibits configuration-dependency. Based on experimental observations, an assumption is taken to simplify the transfer function from the motor input to linkage vibration. Based on this simplification, a dual-mode modal controller is designed and implemented. Experimental results show that linkage vibration is reduced by over 50%

In order to improve trajectory tracking accuracy for a three degree-of-freedom (DOF) planar parallel robot, in this paper, we develop a new control approach based on adaptive control with the use of the so-called synchronization error. Similar to the contour error proposed for machine tools, the defined synchronization error represents the degree of coordination amongst the active joints in the parallel robot, which is substantially different from the traditional tracking errors. By using the synchronization error, all active joints in the parallel robot are controlled to move in a synchronous manner so that the trajectory tracking accuracy of the robot end-effector is substantially improved. In addition, with the use of adaptive control, the synchronization error and the pose error of the platform are guaranteed to converge to zero simultaneously, while uncertain parameters in the system dynamic model are guaranteed to converge to their true values. Experiments conducted on the planar parallel robot verify the above claims and evaluate performance of the proposed control approach, compared with conventional PID control

Modal control algorithms have been widely used in suppressing structural vibration, where vibration characteristics are linear and constant. This paper presents experimental work demonstrating the application of modal control to closed-loop mechanisms, where flexible deformation is coupled with nonlinear rigid body motion. A PRR experimental planar parallel robot is used as the test platform. This lightweight planar parallel manipulator is designed to improve operational speed of "pick-and-place" processes and implement a "smart parallel manipulator" through the integration of a parallel mechanism architecture and active control of linkage vibration using lead zirconate titanate (PZT) transducers. Boundary conditions and mode shapes of intermediate linkage are not conventional due to the fact that the linkages undergo constrained rigid body motion. Experimental modal analysis (EMA) is used to determine the boundary conditions of flexible linkages. However, it is observed that linkage vibration exhibits configuration-dependency. Based on experimental observations, an assumption is taken to simplify the transfer function from the motor input to linkage vibration. Based on this simplification, a modal controller is designed and implemented. Experimental results demonstrate dramatic linkage vibration reduction

High-precision motion of parallel manipulators depends not only on the position accuracy of each actuator, but also on the position synchronization of all actuators. This paper presents a simple synchronized control algorithm for the setpoint position control of parallel manipulators, by incorporating cross-coupling technology into a common proportional-derivative (PD) control architecture. An integrated controller is developed, consisting of a PD control and a saturated proportional-integral (S-PI) control with feedback of the differential position errors amongst actuators (defined as the synchronization errors). The controller can stabilize the motion of each actuator, and meanwhile synchronize all actuators' motions so that both position and synchronization errors converge to zero. The control algorithm does not use the modeling parameters in the controller formulation, and thus permits easy implementation in practice. It is proved that the proposed method can guarantee global asymptotical stability of the system. Experiments conducted on a planar three-degree-of-freedom parallel manipulator demonstrate the effectiveness of the proposed approach.

This paper presents a novel method for the scheduling and control of flexible manufacturing cells (FMCs). The approach employs automata, augmented by time labels proposed herein, for the modeling of machines, transportation devices, buffers, precedence constraints, and part routes. Ramadge-Wonham's supervisory-control theory is then used to synthesize a deadlock-free controller that is also capable of keeping track of time. For a given set of parts to be processed by the cell, A/sup */ search algorithm is subsequently employed using a proposed heuristic function. Three different production configurations are considered: Case 1) each part has a unique route; Case 2) parts may have multiple routes, but same devices in each route; and Case 3) parts may have multiple routes with different devices. The proposed approach yields optimal deadlock-free schedules for the first two cases. For Case 3, our simulations have yielded effective solutions but in practice, optimal deadlock-free schedules may not be obtainable without sacrificing computational time efficiency. One such nontime-efficient method is included in this paper. The proposed approach is illustrated through three typical manufacturing-cell simulation examples; the first adopted from a Petri-net-based scheduling paper, the second adopted from a mathematical-programming-based scheduling paper, and the third, a new example that deals with a more complex FMC scenario where parts have multiple routes for their production. These and other simulations clearly demonstrate the effectiveness of the proposed automata-based scheduling methodology.

In this paper, we develop a new control method for a P-R-R type planar parallel manipulator, termed adaptive synchronized (A-S) control. The novelty of the proposed A-S control, a combination of synchronized control and adaptive control, is in the application of synchronized control to a planar parallel manipulator. To improve trajectory control, based on the kinematics of the planar parallel manipulator, we design a synchronization feedback control signal, termed the synchronization error, which represents the degree of coordination amongst the actuated joints. Employment of the synchronization error is shown to substantially reduce the pose error of the moving platform of the planar parallel manipulator during trajectory tracking. An adaptive controller is also used to estimate uncertain dynamic parameters of the manipulator. Under the assumption of persistent excitation, the proposed A-S control algorithm is theoretically proved to simultaneously guarantee the convergence of tracking errors and the synchronization error. Moreover, the estimated unknown parameters are guaranteed to converge to their true values as well. Finally, experiments are conducted to verify these claims and evaluate the performance of the proposed controller. Experimental results show that A-S control yields good trajectory tracking performance.

A simple synchronized control algorithm is proposed, by incorporating cross-coupling technology into a common PD control architecture, for control of parallel manipulators. A saturated proportional (S-P) control and a linear proportional derivative (PD) control plus gravity force compensation is implemented for synchronization and position control, respectively. The proposed control law is easy to implement and is able to stabilize motion of each actuator while synchronizing all actuators’ motions, so that differential position errors amongst actuators converge to zero. It is shown that the proposed method guarantees global asymptotical stability of the closed-loop system. Experiments performed on a three-DOF parallel manipulator demonstrate the effectiveness of the proposed approach.

A new lightweight planar parallel platform aims to greatly improve operational speed of electronic manufacturing process and to realize a “smart parallel platform” through the integration of a parallel mechanism architecture, new sensing techniques and active vibration control using Lead Zirconate Titanate (PZT) transducers. Experiments using an accelerometer, an impact hammer, and PZT transducers are performed to accomplish Experimental Modal Analysis (EMA), which has not been thoroughly investigated on flexible mechanisms. Through EMA, vibration characteristics of flexible linkages, when the platform is stationary and in motion, are modeled, with configuration-dependent vibration observed. Based on assumptions made in EMA, a modal control strategy using constant mode shapes is designed to control configuration-dependent linkage vibration. Active vibration control experimental results, showing a 40%-50% reduction of vibration amplitudes, validate the effectiveness of the control strategy. This control strategy will be generalized to achieve configuration-dependent vibration control, hence realizing a true smart parallel platform.

In this paper, the dynamic modelling and control design of a planar parallel manipulator used as a pick-and-place machine, is addressed. First, in a departure from standard modelling techniques utilized for planar parallel mechanisms, it is demonstrated that since the translational axes of the manipulator are driven by DC motors through industry standard ball screws, the nonlinear dynamics and coupling effects of the nonlinear dynamics of the manipulator are greatly reduced by a very large effective gear ratio factor, in this case, 1.097 xlO6. The dynamics of the driving motors thus become the dominant dynamics in the system. Hence, the dynamics of the entire system can be approximated as a set of three identical linear dynamic equations, each of which represents the dynamics of one kinematic chain, with constraints representing the coupling of these axes. Then a robust closed-loop controller designed with a Convex Integrated Design (CID) method is determined, such that multiple closed-loop performance specifications, together with a robustness specification, are simultaneously satisfied. The robustness of the closed-loop controller thus guarantees that the controller, although determined based on a simplified linear model, performs as expected on the practical system, i.e., the manipulator, hence results in satisfactory closed-loop performance. Both simulation and experiments conducted demonstrate that the multiple simultaneous closed-loop performance specifications are satisfied thus validating the simplified modeling strategy and verifying the effectiveness of the control design approach.

A high-speed, high-accuracy linear positioning system with application to electronic manufacturing is designed to satisfy four simultaneous closed-loop specifications by integrated design approach. Utilizing the three-stage convex integrated design (CID) method proposed here, the mechanical structure parameters of the linear positioning system, both the control gains and controller structure of a closed-loop controller, are uniquely determined, so that improved closed-loop system performance beyond that required by the n pre-specified closed-loop performance specifications, is acquired. The CID method is especially useful when the design problem has a small set of design variables. Experimental results show that all four performance specifications are simultaneously satisfied, which verify the effectiveness of this design.

For the design of active vibration control laws for flexible linkages in high-speed mechanisms, dynamic models incorporating the coupling of rigid body motion and flexible motion and the electromechanical coupling of transduction devices and the host linkage are very important. In the first part of this paper, the Lagrange finite element (FE) formulation is used to derive such a dynamic model for a flexible planar linkage with one rotary and two translational degrees of freedom. Linear electromechanical coupling of surface-bonded lead zirconate titanate (PZT) patches with the host linkage is incorporated into the model. In the second part of this paper, this dynamic model is simplified and applied to simulate active vibration control of flexible linkages in a high-speed planar parallel platform based on kineto-elastodynamics assumptions: (1) the influence of flexible motion on rigid body motion is ignored; and (2) joint constraint forces in the flexible linkage case are close to those in the rigid linkage case. Based on these assumptions, the rigid body motion and joint constraint forces are regarded as inputs to flexible dynamics, which is obtained from rigid body simulation. Using the simplified dynamic model, simulation of strain rate feedback control using PZT sensors and actuators is performed. Experimental results are also presented. Both simulation and experimental results show that PZT actuators effectively damp the vibration of the flexible linkages.

In this paper, the design and development of a six degrees-of-freedom (DOF) reconfigurable gripper for implementation of robot based flexible fixtureless assembly (FFA) is described. FFA is a novel technique in which traditional fixtures used, for example, in automotive body assembly industries, are eliminated by the use of several robots, with multifinger grippers that are used to grasp and assemble parts. The objective of this work is to design, build, and test a reconfigurable gripper for the purpose of FFA. The gripper developed in this work is capable of precisely grasping parts of different geometries, rigidly holding the parts in space, and choosing different grasping points while allowing sufficient clearance for welding gun access. The overall design procedure, which is subdivided into the conceptual, configuration and parametric design, includes wrench system determination, kinematic and mechanism design. A novel three-finger gripper is developed. Each finger has two movable joints and two point-contacts with which to grasp the part. Finite element analysis is used to simulate deflections of the gripper components under load to determine critical design parameters. Finally, the workspace and the kinematic model of the reconfigurable gripper have been developed. The gripper was fabricated and tested in grasping experiments using several automotive body panels. The reconfigurable gripper is shown to achieve the objectives set out in this work.

This paper describes the process of bonding a MEMS (Micro-ElectroMechanical System) microgripper to the distal end of a robotic manipulator arm using a molten solder bonding technique. This task is part of ongoing work which involves the development of a general microassembly workstation. The goal of this workstation is to construct 3-D microstructures from MEMS sub-components. The microgrippers bonded using the method described here are 1.5 mm by 0.6 mm in size. The methodology behind the solder bonding approach is presented, along with the design of a custom soldering device referred to as the contact head. The contact head is used as the interface between the robotic manipulator and the microgripper. Experimental results are given in a qualitative discussion. An explanation of the bonding procedure using automated calibration is described, along with pictures from the associated microscopy system, and some scanning electron microscope images.

This paper presents a novel method for the scheduling and control of flexible manufacturing cells (FMCs). The approach employs automata. Augmented by time labels proposed in the paper, for the modeling of machines, transportation devices, buffers, part types, precedence constraints, and part routes. The Ramadge-Wonham's supervisory-control theory is then used to synthesize a controller for the workcell representing its deadlock-free behaviour and capable of keeping track of time. For any given batch of parts to be processed by the FMC, a best-first (BF) graph-search strategy is subsequently employed to determine an optimal, or near-optimal, deadlock-free schedule using a novel heuristic function. The proposed approach is illustrated through a typical FMC example, adopted from [HH Xiong, M Zhou, and RJ Caudil IEE Conf. On Robotics and Automation, Albuquerque, USA, Vol. 3, pp. 2793-2797, 1996] to demonstrate the significant reduction in node generation.

This paper presents the development and demonstration of a robotic gripper for active control of shape and vibration of thin-walled flexible payloads. The gripper is a smart gripper, which is configured with multiple actuated fingers that are comprised of linear actuators with DC motors, and laser proximity sensors to enable control of both shape and vibration of thin-walled flexible payloads. A detailed description of the proof-of-concept smart gripper is presented, followed by an experimental demonstration using an automotive front quarter body panel.

A typical problem in flexible manufacturing cells (FMCs) capable of producing multiple parts through multiple routes is optimal routing, where decisions regarding choosing alternative production routes have to be made at certain system states. This paper presents a novel method for determining deadlock-free decisions that optimize a given performance criterion. The approach employs automata, augmented by time labels, for the modeling of machines, transportation devices, buffers, part types, precedence constraints, and part routes. The Ramadge-Wonham's supervisory-control theory is then used to synthesize a controller for the workcell representing its maximally deadlock-free behavior and one that is capable of keeping track of time. This supervisor is utilized to determine the set of optimal decisions. The proposed approach is illustrated through a typical FMC simulation example.

This paper presents a theoretical framework to simultaneously control both static shape deformation and vibration of a flexible payload grasped by an actuated robotic gripper. The dynamic model of the actuated flexible payload is derived using the component mode synthesis (CMS) method with addition of quasi-static modes. Then, the dynamic component model is used to synthesize a simultaneous static shape and vibration controller. Two-time scale control scheme is pursued taking advantage of the two-time scale behavior between the quasi-static modes and vibration modes, which are employed in the model. Simulation results demonstrate the effectiveness of the proposed control approach to simultaneously correct static deformation and suppress vibration in the payload.

This paper presents an improved method of modeling the linear dynamics of a flexible payload grasped by an actuated robotic gripper. The component mode synthesis (CMS) method is employed to explicitly model the coupling between the payload and actuators, and to reduce the system order. With the addition of quasi-static modes to vibration normal modes and constraint or attachment modes, as pre-selected component modes, a new component mode representation is defined. In this study, it is found that the inclusion of quasi-static modes and attachment modes in the CMS formulation results in increased accuracy for simulation of dynamic responses of flexible payloads subject to both static and dynamic external forces. Numerical examples are presented to demonstrate the effectiveness of the new component mode representation for the given problem.

A coordination scheme for two-manipulator systems is developed by maintaining a certain kinematic relationship between manipulators using motion synchronization. The coordination strategy is to let each manipulator track its desired trajectory while synchronizing its motion with the other manipulator's motion so that the differential position error between two manipulators converge to zero. The proposed synchronized controller for each manipulator incorporates the cross-coupling technology into an adaptive control architecture, by feeding back the position error of each manipulator and the differential position error between two manipulators. The proposed algorithm guarantees asymptotic convergence to zero of both position errors and synchronization error. Implementation of this new coordination scheme, which is in a decentralized architecture, is more straightforward. An experiment on two industrial manipulators demonstrates the effectiveness of the proposed approach.

In this paper, the design and development of a six degree-of-freedom reconfigurable gripper for implementation of robot based flexible fixtureless assembly is described. With the flexible fixtureless assembly technique, the traditional fixtures used in automotive body assembly industries are eliminated by the use of robots with multifinger grippers to grasp and assemble parts. The gripper for this purpose has the capability of precisely grasping different parts, rigidly holding the parts in space, and locating the parts for assembly. A novel three-finger gripper is developed in this work. Each finger has two movable joints and two point-contacts for part grasping. The finite element analysis, used to simulate deflections of the gripper components under load, is presented. The overall performance of the gripper is examined. The gripper was fabricated and tested in grasping experiments using automotive body panels, which verified that the objectives set out for this work are achieved.

Coordination of multirobot systems has received extensive studies in the past decade. The majority of previous approaches require a complex setup of the hybrid position/force-control architecture, and have not fully addressed the coordination problem when the robots are not kinematically constrained but perform a common task. In this paper, we propose to use a new coordination scheme that is more straightforward and easier to implement and is applicable to a wider area. The basic idea of the new coordination strategy is to use the concept of motion synchronization, since the problem of coordinating multiple manipulators is basically the problem of maintaining certain kinematic relationships amongst robots. The key to the success of the new method is to ensure that each manipulator tracks its desired trajectory while synchronizing its motion with other manipulators' motions, so that differential (or synchronization) position errors amongst manipulators converge to zero. The controller, designed by incorporating the cross-coupling technology into an adaptive-control architecture, successfully guarantees asymptotic convergence to zero of both position tracking and synchronization errors simultaneously. Experiments and simulations on multirobot assembly systems demonstrate the effectiveness of the approach.

Addresses the application of a fuzzy logic control system to trajectory tracking control of robot manipulators. In the proposed application, the fuzzy logic control system plays the role of a compensator for the robot system, together with the computed torque control method, to improve trajectory tracking performance of an industrial robot. The proposed control scheme is used to adjust weight parameters of a self-tuning fuzzy logic compensator (SFLC). Experimental results demonstrate the effectiveness of the computed torque and the SFLC scheme to control an industrial CRS Robotics Corporation A460 robot.

This paper presents a planar parallel mechanism which can achieve very rapid motion due to the low inertia of its moving parts and the use of multiple simultaneous specification (MSS) control. The proposed parallel manipulator was designed based on a prismatic-revolute-revolute kinematic structure. The proximal prismatic joints were realized using a linear slider with a ball screw mechanism. All actuators remain stationary resulting in a reduction of the inertia of moving parts. Since coupling terms between multiple chains of the parallel manipulator were significant, the MSS control scheme was implemented to satisfy multiple conflicting closed-loop performance specifications. Simulation and experimental results show that the proposed planar parallel manipulator yields better dynamic performance than a conventional X-Y table and has great potential in application to high speed assembly.

Presents dynamic formulations of a planar parallel manipulator including structural flexibility of several linkages. The equations of motion are formulated using the Lagrangian equations of the first type. To avoid complexity in calculating passive coordinates of the parallel manipulator, we introduce Lagrangian multipliers, which combine with constraint equations representing the geometry of multiple closed loop chains. Then, an approach for active damping using a PZT actuator is described to attenuate structural vibration of the linkage. Attached on the linkage, the PZT actuator produces a bending moment according to a linear velocity feedback control scheme. Overall dynamic behavior of the manipulator, induced from structural flexibility of the linkage, is well illustrated through simulations. This analysis is used to develop the prototype parallel manipulator.

A partial model-based torque control algorithm is developed for AC induction motors. The motor is regarded as a torque source, and a desired torque signal is designed based on the desired position trajectory. A reference quadrature axis current input is further developed to cause the torque to track the designed torque signal, utilizing a torque feedforward plus a PI-type torque feedback control. The method utilizes partial knowledge of the electromechanical dynamic model of the induction motor. Experiments conducted on a commercial AC induction servo system demonstrate the validity of the proposed approach.

A control architecture called "advanced torque control" is developed for robot manipulators driven by induction motors. The motor is regarded as a torque source, and a desired torque signal is designed based on the desired position trajectory. A commanded input to the torque loop is further developed to cause the torque to track the designed torque signal, utilizing a torque feedforward plus a PI-type torque feedback control. The method utilizes partial knowledge of the electromechanical dynamic model of the induction motor. Both position and current signals are controlled in the generation of the input to the torque loop. Experimental results conducted on a commercial AC induction servo system demonstrates that this method exhibits good motion performance.

Describes an approach for the use of smart materials, piezoelectric materials of PVDF and PZT, for control of a single-link flexible manipulator. A combined scheme is developed, which consists of a PD feedback for rigid motion control, and a command voltage applied to the PVDF layer or segmented PZT actuator(s) those are bonded to the surface of the flexible link for vibration damping. The command voltage employs linear velocity feedback (L-type), which makes the scheme easy to implement. Global stability of the system is investigated using a Lyapunov approach. It is the first time system stability under the L-type piezoelectric actuator control, utilizing the concept of a virtual joint model, is shown. Simulation results illustrate that the PZT actuator, considering actuator placement, exhibits better performance in vibration damping over the PVDF actuator.

An adaptive-learning (AL) control scheme is developed for control of robotic systems with model uncertainties. When robots perform repetitive tasks, their operations are decomposed into two modes: the single operational mode and the repetitive operational mode. In the single operational mode, the control is a learning based adaptive control where the parameters of the system are updated by using the information of the previous operation. In the repetitive operational mode, the control is a model-based iterative learning control. The advantage of the AL scheme lies in the ability to improve the transient performance at a high rate of learning convergence as robots repeat their operations. Experimental and simulation results ascertain the effectiveness of the AL scheme in controlling a single and multiple robots with model uncertainties.

This paper verifies that the widely used PD plus a force control scheme is also suitable for controlling a system of two industrial robots manipulating a flexible sheet. It is proven by LaSalle's theorem that under the proposed control law, the desired rigid body motion can be achieved and the vibrations of the sheet at each contact are suppressed simultaneously. The offsets of all static deformations of the sheet with reference to the original positions decay to zero. The internal forces between the payload and the robots are well controlled to avoid any damage to the system. The investigation is based on the decomposition of the payload dynamics into two distinct dynamic subsystems, using a "clamped-free" model. The experiments on two CRS A460 robots manipulating a flexible aluminum sheet confirm these theoretical predictions.

This paper describes a new approach to control a single-link flexible manipulator by using smart actuators. A combined PD feedback for rigid motion control and the distributed piezoelectric polymer (PVDF) actuator for vibration damping is investigated using a Lyapunov approach. The PVDF actuator is designed independently from joint velocity of the manipulator, which allows high speed motions. Most of the current PVDF actuators are A-type schemes depending on measurement of the tip angular velocity of the beam, a signal not readily available. This paper is the first to show that the available linear velocity (at the tip) feedback control, L-type scheme, can also guarantee the stability of the system. The L-type scheme does not lead to control problems caused by mode truncation, and is more efficient to suppress the dominant mode vibration of the beam. Simulation results confirm these theoretical predictions.

In a practical controller design problem, several different performance requirements may be encountered together. The goal is to find a controller such that the multiple design specifications, which technically represent the requirements, can be met simultaneously. Such a control problem is called the multiple simultaneous specification (MSS) design problem. Many controller design approaches are proposed to improve the system performance. However, in the robot control area, there is no single design method that can treat a wide range of specifications simultaneously. This paper is concerned with controller design which solves the MSS problem. In this proposed convex combination method, the compromise solution is obtained by properly combining the existing controllers (or control techniques). The design strategy is straightforward and easily implemented. As an illustration, a robotic system is given as an example, and a set of specifications is simultaneously satisfied with the application of this proposed method.

Robotic manipulation of a general flexible object is an extremely difficult and challenging control problem. This paper shows that under a simple PD position feedback, the position/orientation of a general flexible object handled by two manipulators is able to approach the desired one and at the same time the vibration of each contact is suppressed. We use the "clamped-free" model to decompose the motion of the object into two components, a rigid and a flexible one, which allows us to treat them separately and achieve desired motions with a simple PD scheme. This is proved to work theoretically.

This paper presents an active-vision system for the recognition of 3D objects moving along predictable trajectories. The novelty of this system lies in its unique approach to deal with the problem of moving-object recognition, by integrating object pre-marking, object-trajectory-prediction and time-optimal robot-motion techniques developed in our laboratory. The recognition technique is an extension of our earlier work on static-object recognition. Therein, objects were pre-marked optimally using circular markers, which are utilized during run time for guiding a robot-mounted camera to acquire 2D standard-views for efficient matching purposes. The Kalman-filter based prediction of the object trajectory and the time-optimal movement of the mobile camera for image acquisition are also based on our earlier research results on moving-object interception.

This paper addresses methods of control of a multi-robot system designed for the assembly of flexible sheet metal parts. The work begins with the dynamic modeling of two robots and their sheet metal payloads. Recognizing that the flexible sheet metal states are practically unobservable, since none of the sheet metal parts will be instrumented, a practical control algorithm is proposed based on the rigid body dynamics of the robot and payloads. Stability analysis is carried out using the theory of singularly perturbed dynamic systems to determine the effect of the "rigid" feedback on the flexible system. Experimental results are given to illustrate the dynamic response of the two robot system, while carrying out the assembly operation with the two robots under force control. The experiments are performed with two 6-DOF commercial industrial robots.

This paper examines the computed torque control method applied as an outer torque loop supplying desired torque signals to industrial manipulators with flexible, geared, DC motor driven links executing independent inner joint torque control loops. This paper proposes a new control law that restores the desired closed loop dynamic equations to remove the configuration dependence from the manipulator performance through dynamic feedback linearization of the joint torque control signals. Conventional PID, standard computed torque with joint torque control and the new feedback linearized joint torque controllers are applied to the first three joints of a 6 degrees-of-freedom industrial manipulator. The system performance of the controllers in a standard task is evaluated with experiments on an industrial robot. The results show that both computed torque methods provide substantial tracking performance improvements over a conventional PID controller.

In this paper, a generalized implementation of DES-based supervisory controller methodology, that utilizes recent theoretical advances in conjunction with programmable-logic-controller (PLC) technology, is presented. The two primary advantages of the proposed methodology are: 1) the utilization of limited-size control strategies that can be efficiently generated online, and which are conflict and deadlock free by construction; and 2) the use of PLCs, which are currently the most suitable and widely employed industrial process-control technology. In our proposed methodology, a host personal computer (PC) possesses an online capability for the automatic generation of supervisory-control strategies, and their downloading to a PLC as required. The PLC, in turn, is responsible for monitoring the workcell reacting to events and enforcing device behaviour based on the current control strategy residing in its processor. A supervisory controller developed based on this approach, was successfully implemented for a manufacturing workcell in our laboratory.

This paper describes work performed to model the dynamics of a system consisting of two manipulators bringing sheet metal components of a car-body into contact for welding. A sheet metal payload is first discretized into finite shell elements. The flexible payload dynamics are derived via the Lagrangian formulation and combined with the robot dynamics to form one robot-and-payload system. The system equations are simplified by first ignoring certain of the terms that describe the interaction between the flexible and rigid-body coordinates; and second, by applying Guyan reduction. The model developed is applied to simulate the mating of two halves of a car door under three control methods: PD control with gravity compensation, computed torque control, and master/slave hybrid position/force control.

Many sophisticated robot controls are formulated with nonlinear Jacobian transformations, which are dependent on task geometry, as integral components of the control law. Under laboratory conditions, these controls are typically implemented with these nonlinear, task geometry dependent transformations analytically derived and hard-coded in controller software. A change in task geometry implies a change in controller software, a cumbersome exercise that severely restricts the versatility of such controls. In this brief paper, we propose a simple, yet effective algorithm to circumvent this implementation problem. Using only geometric information of the task, the required nonlinear task geometry dependent Jacobian transformations are approximated numerically. While seemingly trivial, this methodology implies that given new task geometry data, these sophisticated task based controls can be utilized without tedious derivation and coding of controller software, a process that effectively renders such controls of little utility in an industrial setting. To illustrate the algorithm proposed, two examples are given with experimental results of one example to contrast the performance of the proposed algorithm with that using hard-coded transformations.

In this paper, a new design of smooth dynamic feedback is proposed that compensates for kinematic drift in motion of a nonholonomic system. The technique is applied to a space-based free-flying manipulator in order to compensate for attitude drift. The proposed drift compensation technique is based on our design of the smooth time-variant dynamic feedback for stabilization of a nonholonomic system. We also present a classification and analysis of various cases of the kinematic drift. Example and numerical simulation are provided for illustration.>

An approach to uncertainty compensation using a multilayer feedforward neural network in multi-manipulator system control is proposed. The proposed approach is developed by formulating the dynamics of the multi-manipulator system in the constrained motion framework. The error-backpropagation algorithm is employed for neural network learning. The teaching signal for neural network learning is derived by analyzing the stability of the closed-loop system. It is shown that if the neural network learns to generate the proper compensating signal, then the constrained motion of the multi-manipulator system tracks the desired motion asymptotically; as a consequence, the desired forces can be achieved. Computer simulations are conducted to verify the proposed approach.>

Considers the problem of control of generalized contact forces with a manipulator controller that has traditionally been regarded as a noncontact task trajectory controller. The open-loop control of generalized forces, suitable for tasks in which only crude force control is required, is achieved through the manipulation of the generalized position inputs of the robotic manipulator. An algorithm is proposed which determines the appropriate manipulator generalized inputs, the only input signal available to the position controller, in order to generate prescribed generalized force and position trajectories during contact with the robot work environment. During noncontact motion of the manipulator, the robot is operated in the more usual generalized position control mode. The use of such a method to control generalized contact forces, although in an open-loop manner, permits a single control to be utilized for both noncontact and contact tasks. Thus, issues of stability during the transition to and from contact, as well as stability during sustained contact are avoided. Hence, the utilization of a single control for both noncontact and contact phases of a single task is seen to be advantageous. The stability of the robotic manipulator during object contact, implicitly assumed by the proposed control strategy, is established using the theory of singular perturbations. Experimental results of a two-degree-of-freedom direct drive manipulator during contact with a one-degree-of-freedom linear mechanical impedance illustrate the usefulness of the proposed method.>

In this paper, the dynamic behavior of a three-layer feedforward neural network as a uncertainty compensator for robotic control is investigated. The investigation is conducted in the context of the robot trajectory tracking problem, where the neural network (with the error-backpropagation algorithm) is used as a uncertainty compensator in conjunction with the feedback linearization control (i.e. computed torque) and a PD control. Through computer simulation, it is verified that the dynamics of the neural network has a specific pattern when the learning rate is sufficiently small, and that such a specific pattern of weight variation in the neural network represents a sufficient condition for closed-loop system performance improvement.

A control methodology that addresses the problem of control of robotic manipulators during a general class of task that requires the manipulator to make a transition from noncontact motion to contact motion and contact motion to noncontact motion is proposed. During noncontact motion, a control suitable for the noncontact phase of motion is applied; during contact, another control, suitable for contact motion, is applied. These different control schemes are applied to the manipulator in such a way that the overall control is discontinuous in nature. The following closed-loop behavior is achieved: (1) the closed-loop system exhibits global asymptotic stability; (2) asymptotic trajectory tracking of generalized force and position inputs is achieved; and, significantly, (3) upon inadvertent loss of contact by the manipulator, contact is reestablished and generalized forces and positions are again achieved asymptotically. Experimental results, performed on a two-degree-of-freedom direct-drive robot, support the theoretical claims.>

The authors investigate the Lyapunov stability property of a control scheme for the bilateral master-slave teleoperator, first introduced by the authors in 1990. Given the nominal models of the master and slave dynamics, and using an approximate feedback linearization control, based on the earlier work of M.W. Spong and M. Vidyasagar (1987), it is shown that Lyapunov stability can be obtained under the assumption that the deviation of the model from the true system satisfies certain norm inequalities. From these norm inequalities, it is shown that the tracking error (position/velocity and force/torque) is bounded and that sufficient conditions for Lyapunov stability can be achieved. The control scheme is illustrated using the simulation of a 30-degree-of-freedom master-slave teleoperator, and the results are presented.>

This paper evaluates the performance of a new control scheme for bilateral master-slave teleo- perator, introduced earlier, and compares it to pub- lished control methods by using a mathematical cri- terion based on the two-port network model. It is shown that most of the published control laws under investigation do not achieve optimal performance, and the few that do, require perfect estimation (or exact knowledge) of the human operator or master arm dynamics. It is also shown that optimal performance, based on the selected mathematical criterion, can be obtained using the proposed control law: I. INTRODUCTION Teleoperating systems may be viewed as synergistic systems composed of human operators and master-slave manipulators which interact with the physical world. The master, sometimes called hand controller, is used to gen- erate commands (usually position, rate or force commands) to its remotedly located counterpart called slave. In gen- eral, the position/velocity command from the operator is fed forward to the slave, while the reaction force, due to mechanical interaction with the environment, is fed back to the master, providing contact force information to the human operator. By reflecting the measured force back to the master arm, it is said that the teleoperator is controlled bilaterally. One of the main issues in telerobotics is control of master-slave teleoperators. A number of control metho- dologies have been proposed in the teleoperator literature. There are two main methods for controlling master-slave teleoperators. The first, a more traditional approach, uses position and/or velocity error (between the master and slave) to generate a forcehorque command into the master arm. The position/velocity of the master and slave are measured by sensors usually located at the joints. The second method uses direct force feedback, rather than posi- tion and velocity errors, to generate command signals to the master. This method uses a force/torque signal obtained from a sensor, usually located at the end-effector of the slave, to provide the operator with the force feeling of the remotedly located counterpart. The force feedback control is usually considered superior because of the more precise nature of the feedback signal. The two-port model network theory has been exten- sively used for the analysis of circuits in which bi- directional energy flows are present at two distinct pairs of terminals. This method provides a useful linear representa- tion of complex networks which exchange energy. The two-port model theory is also used in bilateral teleoperator systems (1,3-81, with force and velocity sensing at the mas- ter and slave. The interfaces between the human operator and master, and between the environment and slave are ports through which the teleoperator is designed to exchange energy between the operator and environment. Recently, several algorithms have been proposed to

An experimental study of a central- izad/decentralized controller for rigid robotic manipu- lators, which solves the robust servomechanism prob- lem for reference signals which asymptotically become constant, is presented. The controller has two tun- parameters which are tuned, using on-line tun- ing methods, so as to optimize the system's speed of response. Exact knowledge of the manipulator dy- namics is not required to implement the controller; in particular, no detailed knowledge of the payload mass is required, i.e. the controller is highly robust to parameter variations of the robotic system. Accelera- tion feedback is the key feature which produces such robust behaviour in the controller. The proposed controller is implemented on a pla- nar five-bar parallel linkage robot and compared to an independent joint PID controller.

The effect of a major source of manipulator compliance, namely, the elasticity of manipulator joints, on the overall stability of robot manipulators during constrained-motion task execution is examined. The stability of the elastic-joint manipulator during constrained-motion contact is investigated separately for the case of two controls applied to the manipulator. Using results from the theory of singular perturbations, the stability of the robotic system is established with a 'rigid' control law applied. The stability of the robotic system is again established using this technique for the case of a 'rigid' control law with a corrective term applied to compensate for joint flexibility applied. It is theoretically established that the presence of joint elasticity does not lead to a destabilizing effect on the manipulator. Numerical simulation results for a two-degree-of-freedom flexible-joint manipulator during constrained-motion task execution confirm the theoretical results.>

The implementation of two robotic control laws proposed in the robotics literature is presented. The discontinuous control law of J.K. Mills (1990) has been demonstrated experimentally to solve the complex problem of control of manipulators during the transition to and from contact motion. Stable transitions are seen during implementation of this control. Further, the control is seen to have a number of highly desirable features, making it attractive for use in industrial manufacturing settings. The constrained motion controller of N.H. McClamroch and D. Wang (1988) has also been implemented. Tests with this control indicate that its performance is as predicted from theory. It is seen that difficulties encountered in the implementation of the constrained motion control result from the lack of a facility to address the problem of inadvertent loss of contact during experimental runs.>

Robotic manipulator impedance control has been presented as a control methodology with the philosophy that a manipulator control system should be designed, not to track a particular motion or force trajectory alone, but rather to regulate the interaction between force and motion. Regulation of generalized contact force and generalized position of the manipulator has not yet been addressed through the use of impedance controllers, as initially proposed in the literature. In this paper, a method is proposed, using an impedance control not only to regulate the interaction between manipulator generalized force and position, but importantly, to additionally control the generalized contact force and position. An algorithm is proposed which determines the appropriate manipulator generalized position inputs, the only input signal available to an impedance control, in order to generate specific generalized force and position trajectories. This goal is achieved while independently regulating the interaction between manipulator force and position. In this paper, the manipulator is assumed to be a rigid structure in frictionless point contact with a work environment modelled as a general linear mechanical impedance. The stability of the robotic manipulator during object contact, implicitly assumed by the proposed control strategy, is established using the theory of singular perturbations. Experimental results obtained with a two degree of freedom direct drive manipulator during contact with a one degre of freedom linear mechanical impedance illustrate the usefulness of the proposed method.>

The nature of the dynamics of a rigid link n-degree-of-freedom robotic manipulator with an r-axis wrist-mounted force-torque sensor is examined during execution of a compliant motion task. A kinematic model is developed that models the flexure of the force-torque sensor during contact, and the dynamic equations of the robot are derived using the Lagrange formulation. In order to display the two-time-scale nature of the resultant dynamic equations of motion, a transformation of coordinates is made. Through utilization of the fact that the mechanical stiffness of the flexural components of the force-torque sensor is typically quite large, the dynamic equations of motion of the robot manipulator during contact can be rewritten in standard singular perturbation form. Using results associated with the theory of singularity perturbed systems, it is shown analytically that the manipulator system can be controlled without consideration of the high-frequency parasitic dynamics associated with the force-torque sensor.>

Practical tasks may require making transitions from free motion to contact motion and from contact motion to free motion. Recently, a discontinuous control law has been shown to provide stability during these transitions. The theory is extended by modifying the discontinuous control law to account for bounded parameter uncertainty. A facility for testing contact task control laws has been designed and built. The facility consists of a two-degree-of-freedom robot, a machine which stimulates a one-degree-of-freedom spring-mass-damper system, and a microcomputer for controlling both of these in real time. The robot has a horizontal, parallel link construction. High-precision direct-drive motors are used which reduce the backlash and friction associated with gear-driven joints. The spring-mass-damper machine is used to create a contact surface with dynamics that can be specified. Software has been designed to make implementation of a wide variety of control laws possible.>

The stability and control of robotic manipulators during the execution of tasks that require the manipulator to make a transition from noncontact motion to contact motion, or vice versa, are investigated. A dynamic model of the manipulator during noncontact and contact motion is developed. This model includes the effect of the inevitable collision that occurs between the manipulator end effector and the work environment during the transition from noncontact to contact motion. The work environment that the manipulator comes into contact with is modeled as a very still surface. The dynamic model of the robot during this transition is transformed through a nonlinear coordinate transformation into a new set of generalized coordinates in which the form of the dynamics is greatly simplified. A discontinuous control is proposed for the robotic manipulator system. It is shown that with this discontinuous control applied to the system, the closed-loop system can be treated as a generalized dynamical system. Using the theory associated with generalized dynamical systems, it is possible to extend Lyapunov stability analysis to systems with discontinuous controls. The system dynamics is written as a contingent equation to which a set valued control function is applied. Within this mathematical framework, the uniform asymptotic stability in the larger of the closed-loop systems is proved. The controller has several desirable properties, including the ability to return to contact motion if the manipulator end effector inadvertently leaves the surface due to some external disturbance acting on the system.>

A hybrid actuation method for robotic manipulators is proposed. The actuator employs a hybrid combination of DC servomotors and muscle-like bladder actuators. One DC motor-muscle actuator pair is arranged coantagonistically with an identical DC motor-muscle actuator pair to drive a manipulator joint. Through a suitable control applied to the hybrid actuator, independent control of joint torsional stiffness and joint position is made possible. When air pressure is varied in the muscle actuators, the muscle actuator stiffness and length change. In order to only affect the hybrid actuator stiffness with this pressure change. DC servomotors are used to compensate for muscle actuator length changes, hence joint position is unaffected. High-gain servomotor control ensures a response to disturbances due almost solely to the muscle actuator stiffness characteristics. Dynamic equations of motion are developed for a two-joint manipulator with a hybrid actuator used to drive the final link. A control is formulated for this system to achieve, in addition to independent joint torsional stiffness and joint position control, approximately decoupled and linearized dynamics. A numerical simulation of this two-degree-of-freedom system is presented to verify the performance of the closed-loop system.>

A complete dynamic model for a rigid link, n-degree-of-freedom robotic manipulator with an r-axis force-torque sensor during contact with a compliant work environment is developed. A Kinematic model is presented for an r-axis force-torque sensor which treats the sensor as an axis joint, where all r axes intersect at a common point. With the resultant kinematic model, the closed-form equations of motion of the manipulator sensor system are developed using a Lagrangian approach. The work environment is modeled as a mechanical impedance with inertia, damping, and stiffness terms. Using of kinematic transformations between the relevant generalized coordinate frames, the work environment dynamic model and manipulator-sensor dynamics are combined into a single set of equations in first-order vector-matrix form. Finally, for completeness, actuator dynamics are included in the system dynamic model, again with the resultant system dynamics in vector-matrix form. A numerical example of a two-degree-of-freedom robot with a two-axis serves to illustrate the kinematic modeling of a force-torque sensor.>

Trajectory control of a manipulator constrained by the contact of the end-effector with the environment represents an important class of control problems. A method is proposed whereby both contact force exerted by the manipulator, and the position of the end-effector while in contact with the surface are controlled. The controller parameters are derived based on a linearized dynamic model of the manipulator during constrained motion. Hence the method is valid only in a neighborhood about the point of linearization. Additionally, a perfect kinematic model of the contact surface is assumed. The proposed method utilizes the fundamental structure of the dynamic formulation of the manipulator's constrained motion. With this formulation, the trajectory control problem is naturally expressed in terms of the state vector variables of the model of the constrained dynamic system. A detailed numerical example illustrates the proposed method.>

A new method is proposed for the accurate trajectory control of a general n degree of freedom robot manipulator, in the presence of large unknown parameter perturbations. In particular, joint friction and large payload mass perturbations are considered. It is demonstrated that for the particular control scheme considered, the choice of high gain feedback results in a system which gives good performance even in the presence of large unmodelled parameter uncertainties. Sufficient conditions for stability are derived using a local subsystem approach. A numerical simulation of a two degree of freedom manipulator, with a large unknown payload mass is included illustrating the new controller design.