The balance between exploration and exploitation is a key problem for reinforcement learning methods, especially for Q-learning. In this paper, a fidelity-based probabilistic Q-learning (FPQL) approach is presented to naturally solve this problem and applied for learning control of quantum systems. In this approach, fidelity is adopted to help direct the learning process and the probability of each action to be selected at a certain state is updated iteratively along with the learning process, which leads to a natural exploration strategy instead of a pointed one with configured parameters. A probabilistic Q-learning (PQL) algorithm is first presented to demonstrate the basic idea of probabilistic action selection. Then the FPQL algorithm is presented for learning control of quantum systems. Two examples (a spin-1/2 system and a Λ-type atomic system) are demonstrated to test the performance of the FPQL algorithm. The results show that FPQL algorithms attain a better balance between exploration and exploitation, and can also avoid local optimal policies and accelerate the learning process.

An approach of control design using fuzzy estimators (FEs) is proposed for quantum systems with uncertainties. Two types of quantum control problems are considered: 1) control of a pure-state quantum system in the presence of uncertainties and 2) control design of quantum systems with initial mixed states and uncertainties. For the first type of tasks, a partial feedback control scheme with an FE is presented to design controllers. In this scheme, an FE is trained to estimate the quantum state for feedback control of a quantum system, and controlled projective measurement is used to assist in controlling the system. For the second type of quantum control tasks, a probabilistic fuzzy estimator (PFE) is trained to estimate the quantum state for control design of a quantum system with an initial mixed state, and a corresponding control algorithm is proposed to design a control law that drives the system from the mixed state to a target pure state. Two examples of two-spin-1/2systems are also presented and analyzed to demonstrate the process of control design and potential applications of the proposed approach.

This paper investigates a perimeter patrol control problem in which multiple agents travel back and forth defending a border. A decentralized control inspired from the adaptive immune system TH1/TH2 differentiation mechanism has been proposed. The resulting patrol control exhibits robustness under adversarial conditions. This is primarily because the motion period can be arbitrary, which makes it difficult for intruders to find weak spots for attack. Moreover, this immuno-inspired control can avoid collision and automatically recover from agent failure.

A new event-based two degree-of-freedom control method for micro-/nanoscale systems based on differential flatness is presented. Compared with traditional time-based feedback control, the proposed method shows great potential in dealing with measurement noise and unexpected disturbances. Instead of using time to drive the planner, we determine the reference trajectory with an intelligent event directly derived from measurement. Therefore, the proposed method inherits the advantages of both feedback and feedforward. Applications to drug delivery demonstrate the promising robustness of this approach.

The key approaches for machine learning, particularly learning in unknown probabilistic environments, are new representations and computation mechanisms. In this paper, a novel quantum reinforcement learning (QRL) method is proposed by combining quantum theory and reinforcement learning (RL). Inspired by the state superposition principle and quantum parallelism, a framework of a value-updating algorithm is introduced. The state (action) in traditional RL is identified as the eigen state (eigen action) in QRL. The state (action) set can be represented with a quantum superposition state, and the eigen state (eigen action) can be obtained by randomly observing the simulated quantum state according to the collapse postulate of quantum measurement. The probability of the eigen action is determined by the probability amplitude, which is updated in parallel according to rewards. Some related characteristics of QRL such as convergence, optimality, and balancing between exploration and exploitation are also analyzed, which shows that this approach makes a good tradeoff between exploration and exploitation using the probability amplitude and can speedup learning through the quantum parallelism. To evaluate the performance and practicability of QRL, several simulated experiments are given, and the results demonstrate the effectiveness and superiority of the QRL algorithm for some complex problems. This paper is also an effective exploration on the application of quantum computation to artificial intelligence.

In this paper, an incoherent control scheme for accomplishing the state control of a class of quantum systems which have wavefunction-controllable subspaces is proposed. This scheme includes the following two steps: projective measurement on the initial state and learning control in the wavefunction-controllable subspace. The first step probabilistically projects the initial state into the wavefunction-controllable subspace. The probability of success is sensitive to the initial state; however, it can be greatly improved through multiple experiments on several identical initial states even in the case with a small probability of success for an individual measurement. The second step finds a local optimal control sequence via quantum reinforcement learning and drives the controlled system to the objective state through a set of suitable controls. In this strategy, the initial states can be unknown identical states, the quantum measurement is used as an effective control, and the controlled system is not necessarily unitarily controllable. This incoherent control scheme provides an alternative quantum engineering strategy for locally controllable quantum systems.

DNA computation is to use DNA molecules for information storing and processing. Challenges currently faced by DNA computation are (1) lack of theoretical computational models for applications, and (2) high error rate for implementation. This paper attempts to address these problems from genetic coding and mathematical modeling aspects. The proposed genetic coding approach provides a promising alternative to reduce high error rate. The mathematical formulation lays down groundwork for studying theoretical aspects of DNA computation

Robotic networks for Autonomous Underwater Vehicles (AUV’s) are an area where efficient use of the communications channel can increase both the mission time and the amount of information communicated between the AUVs, thus resulting in a more effective mission. The traffic in robotic networks must be managed in such a way that takes into account transmission delay, power efficiency, and traditional network measures such as queue length, throughput, and bandwidth efficiency. System optimization of all these criteria is possible because of the cooperative nature of robotic networks. This paper presents a procedure for con structing optimal controllers that limit and route information flow in a network of AUVs. Particular attention is paid to the objectives of maximizing throughput and bandwidth efficiency while minimizing power consumption and queue length. A receding horizon control approach is adopted and shown to yield piecewise continuous optimal controllers that are unique.

This paper addresses the problems associated with missions using multiple cooperating autonomous underwater vehicles. In such missions, inter-vehicle communication is an important requirement upon which coordination strategy and intelligence decision making functions can be built. This paper presents briefly our current modeling and simulation work on characterizing the vehicle communication performance, together with some of the preliminary results. Even with much simplification of scenario modeling, simulation results exhibit intricate communication characteristics, and they can vary significantly as a function of mission requirements and scenarios.

This paper describes a Genetic Algorithm (GA) based approach for area coverage using micro-robots. The method can be used for controlled drug delivery or tumor treatment using micro-robots. The algorithm aims to find a near-optimal path that covers a given area entirely, except obstacles defined as biological barriers or drug side effect restricted zones. The proposed GA approach is a dynamic online path planning approach, which is able to achieve planning during motion and to response to detected obstacles. Different from point-to-point path planning, the proposed operators for the GA are specially designed for area coverage. Comparisons of the approach with high-level static optimal search algorithms and low-level fixed path planning approaches are also presented. Simulation results are given to show the effectiveness of the approach.

This paper proposes an integrated nano-liter bio-material spotting system. The system can be used for DNA gene-chip microarray, protein microarray and other chemical microarray fabrication. The system features a fixed print head, which significantly reduces spotting position and spot size uniformity variations. The system also features a unique spotting pin and print head alignment system design.

This paper surveys methods for drug delivery and discusses micro/nano-devices for controlled drug delivery. The concept of a micro-robot for controlled drug delivery and micro-needles for Chinese acupuncture are also presented. The proposed micro-needles can reduce pain and provide an alternative for drug delivery to the Chinese acupuncture treatment, these micro-needles are fabricated using standard micro-fabrication techniques.

DNA electronics circuits require an efficient way to accurately position and individually manipulate DNA molecules. The recent development of atomic force microscopy (AFM) seems to be a promising solution. We have recently developed an AFM based augmented reality system. This new system can provide both real-time force feedback and real-time visual feedback during nanomanipulation. We have shown that nano-imprinting and manipulation of nano-particles and nano-rods can be easily performed under assistance of the augmented reality system. In this research, the system's ability is extended to manipulation of DNA molecules. Using a polynomial fitting method, the deformation of DNA molecules is displayed in real time in the augmented reality system during manipulation. Indeed, DNA molecules adopt many different structures including kinks, bends, bulges and distortions. These different structures and inappropriate physical contacts may result in the controversy of DNA conductivity reported over the last decade. The AFM based nanomanipulation system can be used either as a nanolithography tool to make small gap electrodes or a nanomanipulation tool to elongate, deform and cut DNA molecules. The measurement of the conductivity of DNA molecules in their different shapes and structures is a promising method to find conclusive evidences, which verify the electrical conductivity of DNA molecules.

The main objective of this paper is to study the control problems for a multi-robot team moving in hierarchical formation. We formulate the problem as a chart by the four-tier hybrid architecture. We devise rules to form a realizable hierarchical formation. Four control laws are derived to achieve the formation patterns. The proposed control laws are suitable for the on-line, real-time formation tasks. Simulation result shows the efficacy of the approaches.

This paper presents a functional model based object-oriented framework for mechatronic system development. The approach has been applied to conceptual design of a new generation plate making system for DNA micro-array manufacturing automation. The contributions of this paper include the functional model and object-oriented development framework for mechatronic systems, and concepts of cost and design optimization at the system functional model level.

This paper first reviews the DNA microarray deposition technology. Different aspects of factory automation for DNA microarray manufacturing are then discussed. Problems, challenges and suggestions for investing factory automation, design, implementation as well as managing automated fab for DNA microarray manufacturing are presented from an industrial automation viewpoint. The purpose is to propose an idea of systematic investment, development and management of factory automation for DNA microarray manufacturing.

This paper proposes a switching control strategy for nonlinear dynamic systems. No special assumptions are made about the type of nonlinearities of the system, except that such nonlinearities are smooth. Stability proof of the switching control, and an example for using the switching control are given in this paper. The switching control framework has been applied extensively for feedback stabilization of the Pendubot - a two link under-actuated mechanical system [Mingjun Zhang et al., March 2002].

Proposes a sensor-based planar object mapping and handling robot system for DNA micro-array deposition equipment. The contributions of the paper include a sensor-based mapping algorithm for planar objects, and the idea of an efficient planar object handling system for DNA micro-array deposition process. The mapping algorithm gives a robust way for mapping a planar object's physical situation based on simple sensor inputs. The success of the algorithm benefits from its unique mathematical modelling of a physical system by simply changing the viewpoint. The new planar object handling system is obtained by integrating the mapping algorithm, and a revised end-effector for current widely used 3-degree-of-freedom robot manipulator.

A method for controlling robots over the Internet, where communication propagation delays exist, is presented. These delays are potentially destabilizing, and certainly degrade the human operator's intuition and performance. A state space formulation is presented, taking into account the time-varying non-deterministic nature of the control and observation delays. A model of the delay characteristics for the communication medium is also derived. Using the state space framework a general-purpose supervisory architecture is developed, allowing the projection of human "intelligence" to the remote environment via the telerobot. Dynamics of the robotic system, as well as the delay characteristics of the communication medium, become part of the design process. The design criteria of transparency, generality, and safety have been met and successfully tested in an experimental setup between Albuquerque, New Mexico and Washington University's Center for Robotics and Automation.

A new hybrid controller has been applied for feedback stabilization of a Pendubot, whose dynamics show second order nonholonomic properties. Experimental results for different test cases show that the hybrid control outperforms the existing control algorithm. This successful implementation provides a novel alternative for the control of under-actuated mechanical systems. Many such systems are subject to second order non-integrable differential constraints.

Since there are many high-rise condominiums, ladder tracks are used against high-rise building fire. However, their height is much higher than the reachable height of the ladders. On the other hand, there have been some studies of vertically moving robots by means of suckers on walls. However, their climbing speed and reliability are not enough for use in an actual fire fighting operation. Therefore, we propose a vertically moving robot which utilizes a balcony handrail as a step. We designed and made a full-size trial robot. We built up the dynamic model of the robot, and planned desired trajectories for the robot. Furthermore we designed a hybrid control system that consists of nonlinear feedback and a sequential method. We concluded that the vertically moving robot is useful for fire fighting against high-rise condominiums from the results of experiments.

Teleoperation permits humans to maneuver robots from a distance. Thus, a human's ability to intelligently manipulate and inspect can be performed in an otherwise inaccessible environment. As a result, teleoperation is gaining acceptance as a cost-effective way to work in remote, often hazardous environments. However, two significant challenges in designing such systems remain: 1. The communication link between the teleoperator and the telerobot's location is usually bandwidth limited and has time-varying delays. Research shows that even a fraction of a second delay between generating a command and observing the corresponding action can seriously degrade the human operator's intuition. This, in turn, diminishes his or her effectiveness. 2. The human teleoperator must rely on artificial means to gain sensory information from the remote environment. This observation is always incomplete due to current bandwidth and sensor limitations. (Also, remember, it is received in a delayed fashion.) The communication channel also limits the fidelity with which the human teleoperator can intervene in the remote environment. The research presented here addresses human/machine cooperation over a bandwidth-limited communication channel with time-varying delays. This cooperation is crucial for taking advantage of the automation's efficiency and the human operator's intelligence.

This paper presents a novel approach for solving a challenging problem in the intelligent control of robotic manufacturing systems, i.e., the integration of low-level system sensing and control with high-level system behavior and perception. First, a newly developed event-based planning and control method will be introduced. It will then be extended to a robotic manufacturing system via a hybrid system approach. The tasks of a robotic manufacturing system usually consist of multiple segments of robotic actions, which involve both continuous and discrete types of actions. The max-plus algebra model has been proposed to model such a system. Combined with the event-based planning and control methods, both discrete and continuous actions can be planned and controlled based on the max-plus algebra model. More important, the interactions between discrete and continuous actions can be formulated analytically. A typical parts-sorting task in a robotic manufacturing system is used to illustrate the proposed approach. The experimental results clearly demonstrate the advantages of this method.

This paper explores a new method for action synchronization and control of telerobotic systems. The significance of the method is that it can effectively deal with the random time-delay existing in the communication channel, such as the Internet. In addition, the result is independent of human operator. First, a novel non-time referenced action control scheme is introduced. Instead of using time as an action reference, a new sensor-based action reference is developed. As a result, the communication delay will have little impact on the operations, in particular the stability of the system. Furthermore, the stability of the system is not dependent on the model of the human operator in the system. All these will lead to the development of the theoretical foundation on the stable control of robotic systems involving human operators. In this paper, the stability condition of robotic system involving both random communication time delay and unknown dynamics of the operator has been developed. It provides the output stability as well as the state stability.

A new method for controlling telerobots over vast distances, where communication propagation delays exist, is presented. A canonical state space formulation is presented, taking into account the time-varying non-deterministic nature of the control and observation delays. A model of the delay characteristics for the communication medium is also derived. Using the state space framework a general purpose supervisory architecture is developed, allowing the projection of human "intelligence" to the remote environment via the telerobot. Dynamics of the robotic system, as well as the delay characteristics of the communication medium, become part of the design process. The design criteria of transparency, generality, and safety have been met and successfully tested in an experimental setup between Albuquerque, New Mexico and Washington University's Center for Robotics and Automation.

This paper presents a novel approach for solving the challenging problem in intelligent control of manufacturing systems. The proposed max-plus algebra model combined with event-based planning and control provides a mechanism to efficiently integrate task scheduling, sensing, planning and real-time execution so that the task scheduling which usually deals with discrete types of events, as well as action planning which usually deals with continuous events, can be treated systematically in a unified analytical model. The unique feature of this approach is that interactions between discrete and continuous events can be considered in a unified framework. This feature allows the manufacturing system to intelligently cope with unexpected events and uncertainties so that the efficiency and reliability of the task schedule and action plan can increase significantly. A robotic manufacturing system is used to illustrate the proposed approach. The experimental results demonstrate the advantages of the proposed approach.

Explores a new dimension in the area of human/robot cooperation: heterogeneous human/robot cooperations. The heterogeneity of human and robot functions can be characterized by the fact that humans are intelligent while robots are fast, powerful and accurate. The important issue, in light of human/robot heterogeneity, is how to plan and control a robotic operation such that the human and the robot can cooperate in a complementary manner. In the paper, the specific problems to be discussed include developing a perceptive action reference frame to integrate human/robot heterogeneous functions; and modeling, analyzing and designing an integrated human/robot system with respect to the perceptive reference frame. These lead to the development of theoretical results on stability of human/robot cooperations, and special methods for planning and control of human/robot integrated systems. The implementations and experimental results presented demonstrate the advantages of the proposed methods.

This paper discusses the planning and control problems of a robot manipulator for a class of constrained motions. The task under consideration is to control a robot such that a tool grasped by the end-effector of the robot follows a path on an unknown surface with the aid of a single camera vision system. To accomplish the task, we propose a new planning and control strategy based on multisensor fusion. Three different sensors-joint encoders, a wrist force-torque sensor and a vision system with a single camera fixed above a workspace-are employed. We decouple control variables into two sub-spaces: one is for force control and the other for control of constrained motion. We also develop a new scheme by means of sensor fusion to handle the uncertainties in an uncalibrated workspace. The contact surface is assumed to be unknown and the precise position and orientation of the camera with respect to the robot is also unknown.

The analytical and robust integration of low-level system sensing and simple control with high-level system behavior and perception is critical technology in the intelligent control of modern manufacturing systems. The paper presents a novel approach by using a max-plus algebra model and a newly designed event-based planning and control scheme to provide an advanced mechanism to efficiently integrate manufacturing systems, so that task scheduling, which usually deals with discrete events, as well as action planning/control, which usually deals with continuous dynamics, can be treated in a unified analytical model. More importantly, the interactions between discrete event systems and continuous dynamics are achieved so that manufacturing systems are allowed to intelligently cope with unexpected system faults and uncertainties during well-scheduled tasks. A robotic manufacturing system is used to illustrate the proposed approach. The experimental results clearly demonstrate the advantages of the proposed approach.

The real-time planning and control problems of a robot manipulator in an unstructured workspace are considered. Our goal is to control a robot to follow an unknown path. Difficulties arise when the robot is required to work in an unstructured workspace. Based on multisensor fusion, a novel strategy for integrating real-time planning and control is proposed to deal with the uncertainties in the environment. In the proposed scheme, a new hybrid position/force control strategy is utilized to maintain the contact between the robot and the unknown surface, while a new planner is designed to complete the path-following task without a priori knowledge of the path. In the paper, to achieve intelligent robot manipulation in an unstructured workspace, multi-sensor fusion is employed both for force-torque and visual sensors with complementary observed data compared to the traditional fusion schemes with redundant data.

A closed form dynamic model of an autonomous underwater robotic vehicle (URV) with multiple manipulators is derived in this paper. The model components include the vehicle body, the m manipulators with n/sub i/ links for each manipulator. It is a generic model that can cope with any number of manipulators and links. The four major hydrodynamic forces included in this model are: added mass, profile drag, fluid acceleration and buoyancy. The model for an URV with two manipulators attached to it is used as an example to show the performance of the model based control scheme. Tracking motions for both vehicle and the two manipulators are tested to show the effectiveness of the nonlinear feedback control scheme. It successfully decouples and linearizes the complex URV system with manipulators, thus attaining a good coordinated control between the vehicle and the attached manipulators.

The essence of our scheme is to extend our earlier (Ghosh et al., 1996) proposed multi-sensor based observation scheme to a situation where the location of the mobile platform is unknown. A new self-calibrated manipulation scheme is proposed to determine the position and orientation of the part with respect to the coordinate system attached to the base of the robot manipulator. The novelties of the proposed approach can be summarized as (i) multi-sensor fusion scheme based on complementary data for the purpose of part localization, (ii) part tracking and grasping control based on parallel tracking, and (iii) self-calibration of the location of the mobile platform using visual data and feature points on the end-effector. The principle advantages of the proposed scheme are described as follows. (i) It renders possible reconfiguring a manufacturing workcell without recalibrating the robot based coordinate frame. This significantly reduces the setup time of the workcell. (ii) It significantly reduces the requirement on the image processing speed. Experimental study of the proposed scheme has been carried out and the results are reported in this paper.

The main problem that we address in this paper is how a robot manipulator is able to track and grasp a part placed arbitrarily on a moving disc conveyor aided by a single CCD camera and fusing information from encoders placed on the conveyor and also from encoders on the robot manipulator. The important assumption that distinguishes our work from what has been previously reported in the literature is that the position and orientation of the camera and the base frame of the robot is a priori assumed to be unknown and is 'visually calibrated' during the operation of the manipulator. Moreover the part placed on the conveyor is assumed to be nonplanar, i.e. the feature points observed on the part is assumed to be located arbitrarily in R/sup 3/. The novelties of the proposed approach in this paper includes a (i) multisensor fusion scheme based on complementary data for the purpose of part localization, and (ii) self-calibration between the turntable and the robot manipulator using visual data and feature points on the end-effector. The principle advantages of the proposed scheme are the following. (i) It renders possible to reconfigure a manufacturing workcell without recalibrating the relation between the turntable and the robot. This significantly shortens the setup time of the workcell. (ii) It greatly weakens the requirement on the image processing speed.

Presents a paradigm for task scheduling and action planning/control of a robotic system. Based on the proposed max-plus algebra model, a robotic task involving both discrete and continuous actions can be scheduled, planned and controlled in a perceptive reference frame. Therefore, task scheduling, which usually deals with discrete type of events, as well as action planning, which usually deals with continuous events, can be treated systematically in a unified framework. More importantly, the unique feature of this approach is that interactions between discrete and continuous events can be considered in the same framework. As a result the efficiency, robustness and reliability of the task schedule and action plan can increase significantly. A typical assembly task in a dual-robot manufacturing work-cell is used to illustrate the proposed approach. The experimental results clearly demonstrate the advantages of the proposed approach.

This paper presents a modular control architecture for multi-arm telerobotic systems. It provides a platform for the implementation of various robot control and planning schemes, as well as a framework for quickly transferring academic research to real-world applications. The advantage is exemplified by the transfer of an intelligent planning and control scheme for telerobotic systems to Sandia National Laboratories. The intelligent planning and control scheme accepts the human supervisory control and autonomous control into one seamless unit. It combines the strength of both human and autonomous control to achieve a safe, robust, and reliable operation. More importantly, the results presented in this paper demonstrate the importance of developing an efficient mechanism for transferring research and technology developed in the academic environment to the national laboratories and eventually into practical applications. It could be a crucial step towards the cross-development of robotic technology in our university, industrial, and national laboratories. The modular controller presented in this paper surely is an important step in this direction.

In this paper we introduce a new approach to visually manipulate, with the aid of a robot manipulator, a part placed randomly on a rotating turntable. The highlight of our approach is that the camera and the robot end effector are both assumed to be uncalibrated. We only assume that the height of the robot end effector is known. Our approach utilizes virtual rotation of the camera via image processing not previously introduced in the literature. Finally, the tracking scheme is implemented by planning the error and gradually forcing it to zero while maintaining the torque controls within acceptable limits. This way we demonstrate a new visually guided analytical tracking scheme.

In this paper, a long outstanding problem, robust impact control and force regulation with guaranteed stability is solved by adding a novel positive acceleration, feedback to the feedback loop in the direction of impact and force control for robot manipulators working in an unknown environment. A quick and stable transition can be accomplished with a nonzero impact velocity, and the transient force response can be controlled with minimum impact forces and minimum bouncing. In addition, a switching control strategy is designed to guarantee the stability of the impact control. Unexpected bouncing can be eliminated after finite switches. The output force and motion can be regulated simultaneously with high accuracy after contact is established. Rigorous stability analysis based on the Liapunov-like method is presented for the proposed control scheme. The control scheme was implemented and tested on a 6 DOF PUMA 560 robot arm. The experimental comparisons with other force control strategies were given to demonstrate the advantage of the proposed method. The result obtained in this paper can be applied to any system requiring impact control and force regulation.

A new planning and control scheme for multirobot coordination is presented. First, the event-based planning and control theory is introduced. The most important step is the design of an event-based motion reference for the multirobot system. It drives the system to achieve the best possible coordination. Hybrid position/force controllers which are able to perform a large class of tasks are designed based on the combination of general task space with the well-known nonlinear feedback linearization technique. To improve the force control performance, the dynamics of joint motors have been considered in the force control. For a given task, a task projection operator can be found for each robot with the consideration of redundancy management. It projects the feedback linearized model to the actual task space. A distributed computing architecture is proposed to implement this scheme in a parallel computation. The event-based coordination scheme was experimentally implemented and tested for the coordinated control of two 6 DOF PUMA 560 robots with very good results.

A new intelligent fusion scheme for human operator command and autonomous planner/controller in telerobotic system has been developed based on the event-based planning and control theory. It provides a unified model to integrate a human operator control command with action planning and control of autonomous operation. As a result, the telerobotic system can perform tasks which cannot be done by either human operator or autonomous planner/controller alone. This scheme lays down a foundation for planning and control of a general robotic system involving human operators, and provides a natural and efficient way to fuse the human intelligence with the machine intelligence. The scheme is implemented and tested on a PUMA 560 robotic system. The experimental results of obstacle avoidance, and hybrid force/position control with a commanded force are presented.

A new sensor-referenced control method for robot impact control and force regulation with environmental uncertainties is developed in this paper. The robot dynamic model is feedback linearized and decoupled for the free motion mode, the phase transition, mode as well as the constrained motion mode. After introducing an impact model, together with the proposed positive acceleration feedback control scheme, a stable contact can be made with a nonzero impact velocity. During the impact, large impact forces and bouncing can be avoided. Moreover, the output force can be regulated after contact is established. The scheme was implemented and tested on a 6 DOF PUMA 560 robot arm. The comparison with other control methods for robot phase transition is presented. The experimental results clearly demonstrate the advantages of the proposed method.

Discusses the problem of using visual and other sensors in the manipulation of a part by a robotic manipulator in a manufacturing workcell. The authors' emphasis is on the part localization problem involved. The authors introduce a new sensor-fusion approach which fuses sensory information from different sensors at various spatial and temporal scales. Relative spatial information obtained from processing of visual information is mapped to absolute taskspace of the robot through fusing of information from an encoder. Data obtained this way can be superimposed upon data obtained from displacement based vision algorithms at coarser time scales to improve overall reliability. Tracking plans reflecting sensor fusion are proposed. The localization of a part by spatial sensor fusion is experimentally demonstrated to be able to give required fast and accurate part localization.

A new function-based control sharing scheme for robotic systems has been developed based on the event-based planning and control theory. It provides a unified model to integrate a human operator control command with action planning and control of autonomous operation. As a result, the robotic systems can perform tasks which cannot be done by either human operator or autonomous planner/controller alone. This scheme lays down a foundation for planning and control of a general robotic system involving human operators, and provides a natural and efficient way to fuse the human intelligence with the machine intelligence. The scheme is implemented and tested on a PUMA 560 dual-arm system. The experimental results of obstacle avoidance, hybrid force/position control with a commanded force as well as dual-arm coordinated teleoperation are presented.

A multi-sensor based planning and control scheme for robotic manufacturing is presented in this paper. The proposed approach fuses sensory information from various sensors at different temporal and spatial scales in an event-based planning and control scheme. By combining the measurement of an encoder sensor, relative spatial information obtained from processing of visual measurement is mapped to the absolute task-space of the robot, delayed data obtained from a displacement based vision algorithm that represent absolute part position measurement is brought to up to date. A four-step approach to planning and control of a robotic manipulator is discussed. An event-driven tracking and control scheme that is based on multi-sensor information is given. The approach is illustrated by considering a manufacturing workcell where the manipulator is commanded to pick up a part on a disc conveyor under the guidance of computer vision.

Using semi-algebraic description of task level configuration space (TLCS) and calculus of variations, trajectory plans avoiding the otherwise invisible, TLCS obstacles, arising from the kinematic constraints due to the joint limits are synthesized. This computationally tractable trajectory planning consists of shortest and smooth geometric path planning satisfying kinematic constraints, and event-based motion planning. As opposed to the conventional algorithmic approaches to the path planning, this appealing closed form solution of the path and motion planning, and real time control is easily applicable to computer integrated manufacturing applications. The validity of this synthesis along with a real time control is demonstrated through simulation and experiment on a PUMA 560 manipulator.>

A novel event based motion planning method, together with nonlinear feedback algorithm is applied to the controlled transition between unconstrained and constrained motion of a rigid robotic manipulator. The motion planner is driven by a task related scalar motion reference. This reference is a function of the actual manipulator position in the unconstrained directions and a function of the interaction force in the constrained directions. Impact is considered a discrete event, separating different operation modes. The dynamics of the arm is decoupled with respect to a path-relative representation of manipulator outputs. The feedback control law directly servos these outputs. The event based planning and control scheme was experimentally implemented for impact control and compared with impedance control and explicit force control.>

This paper describes a distributed computing and control architecture for experiments in dual-arm sensor-based planning and control. The system comprises of two PUMA 560 robots with state-of-the-art joint-level motion controllers. These controllers have accurate joint position and velocity measurement, an accurate joint calibration system, and six-axis wrist force/torque sensors with optical fiber communication with the controller. A high-power Silicon Graphics workstation, through a shared memory interface with the joint-level controllers, provides a task-level command and control capability. A variety of control schemes can be easily implemented on this system. A servo rate of 1000 Hz, without interpolation, has been achieved for task-level servo schemes with force feedback. The experimental results presented here demonstrate the performance capabilities of the system.>

Based on our newly developed event-based planning and control theory, a new action planning and control scheme for the telerobotic system is presented. It provides a unified model to integrate a human operator control command (supervisory command) with motion planning and control for automatic operation. As a result, supervisory command and autonomous motion planning and control are naturally combined together such that the autonomous motion plan can be adjusted and modified by human operator during its execution. This scheme lays down a foundation for planning and control of a general robotic system involving human operators. The scheme is implemented and tested on a PUMA 560 dual-arm system. A Spaceball is used by the human operator as a command generator to input the supervisory command. The experimental results of both single arm teleoperation and dual-arm coordinated teleoperation are presented.>

Successful execution of a planned task using robotic manipulators, requires that the task satisfies both the kinematic and dynamic abilities of the manipulator. The executability of a task is generally unknown a priori. In this paper, the dynamic ability of a manipulator for motions constrained by the holonomic constraints and by the loading in the object maneuver are considered. Simulation and experimental results for holonomically constrained motions are presented to illustrate the usefulness of the presented formulation in analyzing the task executability and preventing manipulator damage for some tasks, especially, those involving environmental interaction.>

The problem of rigid object handling by multiple robot arms is investigated. The primary goal is to make the object exhibit a prescribed behavior while in contact with a fully known environment. Point contacts are assumed between the object and the arms. The aspect of task definition to achieve decoupling and linearizing control laws is discussed. Control laws are first formulated at the object level to provide decoupled force and position servo loops. It is then used to form control laws for the individual arms. Redundancies exist at the object and arm levels. The object level redundancy is used to achieve secondary goals in object handling. The arm level redundancies are the zero dynamics and can be controlled by redundant inputs. Full use of the available inputs are used to control the system as a whole. Numerical simulations for a dual-arm situation illustrate the validity of the approach.>

Abstruct- This paper presents new robot aim1 planiiiiig and control schemes compatible with sensing based planning and control. These developiiieiits are a central component of robot arm coiitrol intelligence. The sclieines iise pliase space (velocity versus position) representation of robot arm motion in tlie task space, and apply time and energy optimization tecliiiiqiics to deteriiiiuc non-time based trajectories for giveii geometric path in tlie task space. A new pliasc space crror defiiiitioii aiid coiripiitatioii liavc been introtliiced aiid combined witli kiiowii nonlinear fcetlback control law, whicli linearizes and decoiiples the control in tlie task space. Tlic! system stability lias been proveii. Tlie new phase space plaiiiiiiig aiid coiitrol schemes were experiiiieiitally iiiipleiiieiited and tested on the 3-DOF positioii coiitrol of a PUMA 560 robot arm with very good results. I. INTRODIICTION

The discrete time theory that feedback linearizes and output decouples the coupled nonlinear dynamics of the PUMA 560 robot arm is systematically developed. The effect of sampling on the performance of the arm has been studied and shown through experiments. Certain restrictions that are necessary on the sampling time of the system are shown. The results obtained have been experimentally verified. Since a number of researchers have addressed the problem of loss of feedback linearizability under sampling, it was important to develop theories and obtain results in discrete time that take into consideration the effects due to sampling. This problem is addressed.>

The trajectory tracking performance of the nonlinear feedback controller based on differential geometric control theory was experimentally studied. The performance is compared to three other methods: independent joint proportional-derivative (PD) control, independent joint PD control with feedforward torque computations, and the computed torque method. In the nonlinear feedback controller, servo is in the task space. In the other three methods servo is in the joint space. Path planning is in the task space. Circular and straight line paths are considered. A sampling frequency of 200 Hz is used. The effect of sampling frequency on tracking error is studied for the nonlinear feedback control scheme. The performance of these methods was tested on a PUMA 560 arm. The trajectory tracking performance of the robot arm under the nonlinear feedback controller is comparable to that of the computed torque control schemes.>

The authors focus on the investigation of maximum applicable force/torque by multiple robot arms in coordinated motion, i.e., the dynamic workspace analysis of multiple cooperating robot arms. Simple procedures to solve the dynamic workspace analysis problem for multiple cooperating robot arms with or without the internal force/torque constraint are presented. These procedures model the cooperating robot arms as a closed mechanical chain and are based on the theory of linear transformations and the properties of mechanics. For low degree-of-freedom robot arms, the maximum force/torque problem can be solved by geometrically using the proposed procedures, as illustrated by examples.>

A nonlinear feedback robot controller that incorporates the robot manipulator dynamics and the robot joint motor dynamics is proposed. The manipulator dynamics and the motor dynamics are coupled to obtain a third-order-dynamic model, and differential geometry control theory is applied to produce a linearized and decoupled robot controller. The derived robot controller operates in the robot task space, thus eliminating the need for decomposition of motion commands into robot joint space commands. Computer simulations are performed to verify the feasibility of the proposed robot controller. The controller is further experimentally evaluated on the PUMA 560 robot arm. The experiments show that the proposed controller produces good trajectory tracking performances and is robust in the presence of model inaccuracies. Compared with a nonlinear feedback robot controller based on the manipulator dynamics only, the proposed robot controller yields conspicuously improved performance.>

The nonlinear feedback method of robot control has been experimentally implemented on two PUMA 560 robot arms. The feasibility of the proposed controller, which is already known to be viable, is stressed. The servomechanism operates in task space, and the nonlinear feedback takes care of the necessary transformations to compute the necessary joint currents. A discussion is presented of the implementation with details of the experiments performed. The performance of the controller is encouraging but was limited to 100 Hz sampling frequency and to derived velocity information at the time of the experimentation. The setup of the lab, the software aspects, results, and the control hardware architecture that has recently been implemented are discussed.>

Hamilton's principle is used to derive the dynamic model for a large class of flexible robot arms. The resultant dynamic model consists of a system of partial differential-integral equations and the dynamic boundary conditions associated with it. Some properties of the model are observed and its application to control is discussed. This model represents an infinite-dimensional nonlinear dynamic system and yet can be turned into a finite-dimensional system that could be obtained by modal expansion, if it is desired. This provides more flexibility for control purposes as well as for the analysis of the system.>

Based on a differential geometric feedback linearization technique for nonlinear time-varying systems, a dynamic force control method for robot arms is developed. It uses active force-moment measurements at the robot wrist. The controller design fully incorporates the robot-arm dynamics and is so general that it can be reduced to pure position control, hybrid position/force control, and pure force control. The controller design is independent of the tasks to be performed. Computer simulations show that the controller improves the position error by a factor of ten in cases in which position errors generate force measurements. A theorem on linearization of time-varying system is also presented.>

In this paper the dynamic equation describing two cooperating robot arms which simultaneously are working on the same object is established by considering the whole system as a closed kinematic chain. This formulation has the advantage of automatically handling the coordination and load distribution between the robot arms through the dynamic equation. It can also be generalized to a multi-robot arm system. The control problem is solved as a design problem for a linear system by first applying a nonlinear feedback and nonlinear coordinate transformation to linearize the nonlinear dynamic equation, and then employing the linear optimal control theory to design a robust controller in the task space. This new dynamic control method establishes a direct control response to task space commands facilitating dynamics based task encoding in robotics.

A new control method is presented for the coordinated control of two robot arms based on exact system linearization by appropriate non-linear feedback. The two arms are assumed to be working on the same object simultaneously. The control method uses a dynamic coordinator acting on relative position and velocity error and/or on relative force-torque errors between the two arms. Simulation results are presented for the coordinated control of two PUMA 560 robot arms using an optimal dynamic coordinator. For completeness, the paper also presents (i) a new revised set of inertia parameters for all six links of PUMA 560 robot arm, (ii) the form of nonlinear feedback and diffeomorphic state transformation which linearizes the control of PUMA 560 working in all six degrees of freedom, and (iii) the reduced numeric form of all dynamic equations for PUMA 560 working in three (positioning) degrees of freedom.

A new dynamic control method is discussed for robot arms. It is based on nonlinear feedback and diffeomorphic (state) transformation which externally (or exactly) linearizes the whole system and provides simultaneous output decoupling. The non-linear feedback is augmented with optimal error correcting controller, which operates on the task error level. Several computer simulations were performed to evaluate the performance of the new dynamic control method. The simulation results are discussed in detail. The simulations show that the new dynamic control method provides very good tracking performance even in cases when the dynamic model parameters of the robot arm are only known with 10-30% error. The functional characteristics of the nonlinear feedback law and optimal error correcting controller are discussed briefly.