LIMITATIONS in current robot compliance controls motivated

Size: px
Start display at page:

Download "LIMITATIONS in current robot compliance controls motivated"

Transcription

1 586 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 Analysis and Implementation of a Neuromuscular-Like Control for Robotic Compliance Chi-haur Wu, Senior Member, IEEE, Kao-Shing Hwang, Member, IEEE, and Shih-Lang Chang Abstract In comparison with robot manipulators, primate limbs excel robots in facile movements requiring compliance control. Based on this fact, this paper will extend our findings in modeling the muscle-reflex mechanism of primate limbs to robotic control. After some salient properties of the neuromuscular system were identified, a neuromuscular-like model that can accurately emulate different involuntary and voluntary movements was developed. To link the findings from the biological system to robotic control, the developed neuromuscularlike controller was implemented on a PUMA 560 robot. The experimental results demonstrated that the emulated spindlereflex model in the neuromuscular-like controller acts as an impedance to any changing displacement and will comply and enhance the needed compliant forces or torques for the changing motion. Due to this force-enhancement property, no external force sensor is required for sensing force feedback in this control. The capability in performing various free and constrained movements demonstrated that a neuromuscular-like control is very useful for robotic applications requiring adaptation. Index Terms Biological control systems, biological system modeling, control systems, force control, man machine systems, nonlinear systems, robots, system analysis and design. I. INTRODUCTION LIMITATIONS in current robot compliance controls motivated the study of this paper. In robot motion maneuvers, there are two types of motion, unconstrained and constrained. Unconstrained motion requires only position control, which is the motion that most of industrial robots are allowed. The constrained motion, such as assembling parts, requires compliance control that combines both position control and force control [1], [2]. In recognizing the importance of compliance control for advanced robotic applications, many approaches were proposed. One approach selected free joints for the motion and controled these free joints with forces and other joints with positions [3]. Another approach controled both positions and contact forces generated at the robot wrist to provide compliance [4]. Other approaches have used force feedback strategies, such as accommodation force control [5] and active stiffness control [6]. Advancing the concept of stiffness control to the control of mechanical impedance about the limb joint by coactivation of antagonist muscles, an impedance control was also proposed for robot compliance [7]. Manuscript received January 2, 1996; revised March 4, Recommened by Associate Editor, B. Espiau. C. Wu and S.-L. Chang are with the Department of Electrical and Computer Engineering, Northwestern University, Evanston, IL USA. K.-S. Hwang is with the Department of Electrical Engineering, National Chung-Cheng University, Chia-Yi, Taiwan. Publisher Item Identifier S (97) Although many approaches were proposed, the performance of industrial robots is still crude in providing compliance. This deficiency remains one of the major problems limiting the scope of robotic applications. It is widely recognized that the primate limb has much superior performance for delicate, skillful maneuvers, especially in adaptation to different loads and capability to execute compliant tasks. It seems that a good compliance control for robotic applications should emulate the compliant capability of the neuromuscular system. However, only few developed schemes for robotic compliance control take advantage of findings in biological limb research. Hogan s impedance control relates to the spring-like interface between limb and environment. Modeling the spring-like behavior of the neuromuscular system, while the limb is in contact with the environment, allows successful impedance control to provide compliance for a robot. The concept of flexion and extension in muscular system was also adopted by Jacobsen et al. to control manipulator links with two tendon-driven actuators [8]. Their control algorithms in position and force control showed good results, experimentally. The success of these efforts indicates that a better design of robotic compliance control may benefit from modeling the mechanisms of biological limbs. According to an extensive body of experimental evidence [9] [11], a nonlinear property extracted from force-velocity data shows that force response is proportional to a low fractional power, 0.17, of muscle s velocity. From our initial study, we found that this unusual nonlinear damping property acts as a control function to regulate different limb movements [12]. Intrigued by this property, we further studied the characteristics of the neuromuscular system. Although the nonlinear dynamics of the muscle-reflex system have been studied previously [13] [16], the neurophysiological models are too complicated to be simulated or to be used for robotic control. To develop a neuromuscular-like control model for practical applications, the mechanisms of the muscle-reflex system were studied. According to the biological model, the muscle-reflex mechanisms form a feedback system called the motor servo [9], [15], consisting of a muscle, its spindle receptors, and the corresponding reflex pathways back to the muscle. This neuromuscular system mediates the stretch and unloading reflex of the muscle by the feedback. To replicate the musclereflex system, Gielen and Houk proposed a model of the motor servo that incorporated a nonlinear description of the behavior of muscle spindle receptors, a delay in the reflex loop via the spinal cord, and a nonlinear model describing muscle /97$ IEEE

2 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 587 Fig. 1. Muscle-reflex model. mechanical properties [15]. To simplify the biological model proposed by Gielen and Houk [15] for practical applications, a muscle-reflex model shown in Fig. 1, consists of a spindlelike model and a simple muscle stiffness mechanism was developed first to emulate various involuntary movements [17]. The parameters in the model were optimized through fitting the experimental data. This model will be explained in Section II. After further fitting the experimental data of various voluntary movements. a neuromuscular-like model was developed for emulating limb movements [18], [19]. A very important property was found that the control based on the neuromuscular-like model can provide the needed reflex and compliance for moving the limb. The emulated spindlelike model in the neuromuscular-like model will react reflexly to any changing displacement. In other words, when it senses a small change in limb position, it will reflexly enhance the muscle force to move the limb. Due to this property, the neuromuscular-like model can emulate various voluntary movements (free movements) through motor commands and adapt to the transition from voluntary movements to involuntary movements (constrained movements) without using a force sensor. This capability is essential in designing a compliance control for robots. Current industrial robots are poor in compliant control and adaptability, a neuromuscularlike control seems to be a reasonable approach to improve that. In this paper, the practicality of modeling the muscle-reflex system for actual robotic applications will be demonstrated with experiments on a PUMA 560 robot. II. A MODEL OF NEUROMUSCULAR-LIKE CONTROL The proposed neuromuscular-like control is modeled from the physiological properties of the primate muscle-reflex mechanisms. After studying the properties of the muscle-reflex mechanisms, a muscle-reflex model shown in Fig. 1. was developed first [17]. This muscle-reflex model can emulate the response of extensor or flexor for the involuntary movement. However, to emulate voluntary movements, a more complete model is required for performing the operations of both extensor and flexor. In other words, voluntary movements require torques in two directions (flexion and extension), and therefore two separate muscles. Involuntary movements are driven by external forces, but voluntary movements are dictated by the CNS (central nervous system) through motor commands as well as by external forces. Based on these basic distinctions between voluntary and involuntary movements, a neuromuscular-like model, depicted in Fig. 2, consists of a pair of agonist and antagonist muscles for flexion and extension was proposed [18], [19]. The proposed neuromuscular-like model was originally modeled from two muscles producing opposing torques on a hand. The hand rotates about the wrist with a single degree of freedom (DOF). In Fig. 2 and this paper, the subscripts ago and ant represent agonist and antagonist, respectively. The neuromuscular-like model proposed in Fig. 2 has two muscles. Each muscle is emulated by a muscle-reflex model that consists of two major parts: spindle-like mechanism emulating the reflex property of the biological muscular system for absorbing impacting forces, and muscle-stiffness mechanism emulating muscle stiffness for tracking various movements. The analogy of this muscle-reflex model is depicted in Fig. 1. The model has two possible inputs, motor command and external disturbing force. The load movement is represented as. Whenever there is a load movement, the muscle length will vary with load position [9]. This equilibrium movement at the muscle is represented as, where is the initial position of the load sensed at the muscle. The reflex signal from the spindle-like model, scaled through a reflex gain coefficient, is combined with the motor command to produce a reflex-induced command for generating muscle force. However, to simulate the smoothing of commands by motor neuron pools such that a bell-shaped velocity profile for limb movements can be induced, a lowpass filter (LPF) must also be included in the model. Based on the experimental data, a time constant of 30 ms is extracted for this filter effect [15]. The linear feedback gain coefficient shown in Figs. 1 and Fig. 2, represents the effect of muscle length-tension, which represents any spring-like behavior in the muscle-reflex system [9]. This effect describes that any change in muscle length will produce a muscle force through muscle stiffness. In sum, both spindle and muscle stiffness mechanisms will induce muscle forces. The resultant muscle force combined with the disturbing force at the load will then move the load at the hand with a movement and cause an equilibrium movement of at the muscle. If the hand is modeled by an inertial load with a damping constant, the dynamic system under the neuromuscular-like

3 588 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 Fig. 2. Neuromuscular-like two muscle model. control can be represented as where and represent the load velocity and acceleration, respectively. The muscle force is then produced by two muscle-reflex models shown in Fig. 2, representing an agonist and an antagonist. Each muscle-reflex model consists of a simple muscle stiffness mechanism, driven by the sum of a voluntary command and a reflex signal produced by a spindlelike model. Muscle-stiffness mechanism is assumed to be spring-like, with a spring constant. With this stiffness, muscle force is generated as follows: In the above equation, and represent the length-tension effect of the agonist and antagonist to the muscle force, respectively; and and represent the equilibrium movements sensed at agonist and antagonist, respectively. The equilibrium movement at each muscle is obtained from the difference between its initial position, or, and the sensed load movement as depicted in Fig. 2. As for and, they represent lumped neural inputs to the agonist and antagonist muscle, respectively. These two lumped inputs are the outputs of two LPF s represented in Fig. 2. The purpose of these LPF s is to induce the bellshaped velocity profile for the voluntary movement. (It might represent smoothing of motor commands by motor neuron pools.) This effect is represented as follows: (1) (2) LPF (3) LPF (4) where and are the neural signals before filtering, and LPF represents a LPF with a time constant, which is about s based on the experimental data [15]. In (3) and (4), each is the sum of voluntary motor command and reflex signal generated from the spindle-like model. This relationship at the agonist and antagonist can be represented, respectively, as (5) (6) where and are the reflex signals generated from the spindle-like models of the agonist and antagonist, respectively; and is the reflex gain coefficient. In order to model the spindle-like mechanism in Fig. 2, physiological properties of the spindle mechanism have to be known. The physiological studies indicated that the motor servo of neuromuscular system has two salient nonlinear dynamic features, a stiffness enhancement at small signals and a fractional velocity-dependent viscosity [13] [16]. The stiffness of the neuromuscular system, which is not constant over the stretching range, has more effect on the initial elastic force response, a property which is referred to as the short-range enhancement. The velocity-dependent viscous force response, however, shows a nonlinear viscosity property that has been postulated to damp limb movements when variable loads are involved. In sum, spindle receptors of the muscle-reflex system will sense both muscle length and the rate of change of muscle length and produce a firing rate reflecting both measurements. Therefore, the simulated spindle-like model in our model must capture the physiological properties of the described nonlinear damping effect multi-

4 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 589 plied by the short-range elasticity enhancement. After fitting recorded experimental data in our previous research [17], a spindle-like model was developed for each muscle as follows: where and are the internal position, velocity, and bias position of the spindle-like model for the agonist (the antagonist counterparts are with a subscript of ant ) and is the reflex stiffness and is a scaled damping coefficient for each spindle-like model. In this spindle-like model, the short-range elasticity enhancement for moving either direction (positive or negative) was fitted through experimental data and modeled as for the agonist muscle; similar effect was also modeled for the antagonist. According to (7) and (8), any change of load position will cause an equilibrium movement at each muscle, represented as and. Then these changes will induce reflex signals and at the corresponding muscles by the spindle-like model. And these reflex signals scaled by a reflex gain will then produce muscle forces to respond to the change. With property of the nonlinear damping effect multiplied by the short-range elasticity enhancement, this spindle-like model will provide a property of force enhancement and acts as a control function regulating the limb motion in such a manner that the neuromuscular-like system will enhance the limb motion when high velocity is needed; conversely, it will also help stop limb motion promptly when velocity decreases to the low-velocity range. In other words, the spindle-like mechanism is capable of making automatic corrections in muscle forces to respond to external disturbance forces, rapidly. Therefore, it has the capability in absorbing any impacting force that causes load displacements by complying muscle forces or torques to the motion. To identify the parameters of the neuromuscular-like model, different sets of experimental data collected from different involuntary and voluntary movements were used as the templets for fitting the proposed model. These experimental data were provided by the Laboratory of Professor J. Houk, Department of Physiology, Northwestern University, Evanston, IL. In our simulated limb wrist model, was set to one so that represents the muscle mechanical stiffness describing the linear effect of muscle length-tension. As the involuntary movements involve no motor commands, and were both set to zero. After minimizing the mean-square errors between the simulated results and the experimental data, a set of parameters were obtained as follows: m/n, N/m, N/m, N/m (s/m) m, and RC s. The initial study of involuntary movements demonstrated that the neuromuscular-like model has appealing features for controlling constrained movements [17]. Then, the study of voluntary movements further proved that the proposed model can perform both free and constrained movements as required by robotic compliance control [19]. For the purpose of enhancing the capability of robots in compliance, these (7) (8) encouraging results from our initial studies prompt the research done in this paper. III. IMPLEMENTATION OF THE NEUROMUSCULAR-LIKE CONTROL ON ROBOT In the research of limb control, limb movements are usually assumed to be under the equilibrium trajectory control [7], [24]. Under such a hypothesis, the brain sends equilibrium trajectory to the limb without worrying about the low-level control problems. This theory makes maximum use of viscoelastic properties of the muscle-reflex system and treats both free movements and constrained movements coherently. In other words, a lot of complicated computations can be avoided if the low-level control system can adapt to different task conditions. From this hypothesis, the neuromuscular-like control seems to be a good choice for low-level control. Therefore, the main goal of this paper is to link our findings in modeling the biological muscle-reflex system to the development of a control method that can aid the performance of an industrial robot. The developed controller will be implemented on a PUMA 560 robot. A. Motor Commands Patterns Under the hypothesis of equilibrium trajectory control, limb movements can be partly modeled by velocity profile and force profile. According to a bell-shape velocity profile, minimum-jerk criterion was proposed to emulate human arm movements as straight lines [21]. Based on minimum-jerk criterion, a movement trajectory can be described by a fifthorder polynomial. In contrast with minimum-jerk planning, Uno et al. proposed a minimum-torque method, which can emulate certain limb movements that are not straight lines. To relate planning with dynamics, Hollerbach and Atkeson pointed out that there is a fundamental, time-scaling relationship between manipulator s velocity and its dynamics that allows trajectory planning and inverse dynamics to be coupled exactly and efficiently [22]. To move a multijoint limb, Flash also modeled the arm movement based on the equilibrium trajectory control [23], in which each equilibrium point is generated by a spring-like elastic energy function [24] with the directional position-dependence stiffness and viscosity. To determine a set of motor commands for a multiplejoint robot, we must decide the domain for planning these commands. In investigating whether planning variables of human arm movements are at the Cartesian domain or joint domain, a staggered joint interpolation was proposed [22], in which a time delay is introduced to one joint to slow down that joint. This method could approximate straight lines in the Cartesian space. They concluded that this planning may best describe human movements. In our applications, we will adopt their concept of staggered planning to plan motor commands representing position commands. Given a starting position and a final position, motor commands for this movement will be mathematically planned by the following equation: (9)

5 590 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 Fig. 3. One-muscle implementation on joint i of a PUMA 560 robot. where represents a time delay if the staggered planning is used. This formulation has the advantage of representing most of command types we are going to investigate. For example, if is a constant, then the command type is step; however, if is a first-order function, then it is ramp. If a minimum-jerk planning is required, then will be a fifth-order function of time. The minimum-jerk planning has been used successfully in emulating straight-line human movements for a two-joint limb by other researchers [21], [25]. To provide a smooth planning for an industrial robot such as PUMA 560, the planning of minimum-jerk in (9) will be adopted for our experiments. The concept of staggering strategy will also be applied in our planning if the geometrical shape of a path needs to be maintained. Based on those patterns, motor commands representing positions will be planned by (9) and sent to the neuromuscular-like controller. B. Neuromuscular-Like Control for Robot Joints A six-joint PUMA 560 robot in our laboratory is used to implement the proposed neuromuscular-like controller. The ideal neuromuscular-like model, shown in Fig. 2, consists of two muscles producing opposing forces on a joint. However, to implement this controller on PUMA 560 robot, two-muscle implementation is not feasible because there is only one motor at each joint. To accommodate this model for the PUMA robot, the single-muscle model shown in Fig. 1 is implemented as Fig. 3 to drive each joint of PUMA robot. Since this implementation has only one muscle, there are no coactivation of two muscles in this model. The mechanisms of this singlemuscle model acting on a joint will be explained in the following. In general, the dynamic system of an degree-of-freedom robot is represented by the following equation: (10) where and represent joint positions, velocities, and accelerations of all joints, respectively; and represent the inertia matrix, the Coriolis and centripetal force vector, and the gravity force vector, respectively; and represents applied joint torque/force vector. In our earlier study, we found that a system controlled by the nonlinear damping in the neuromuscular-like model can adapt to different inertial loads [12]. Based on this property, the neuromuscular-like control seems to be a good candidate for controlling a robot with changing configurations. To utilize the neuromuscular-like controller for controlling this nonlinear robot, the controller will generate the required muscle force and apply it to each joint. In other words, the applied torque/force will be equivalent to the muscle force computed from the neuromuscular-like controller for all joints. To design a neuromuscular-like controller for each robot joint, we will model the dynamic system of joint of a robot as (11) In this equation, and represent velocity and acceleration of joint, respectively. represents the joint inertial load varied with the changing robot configurations,, and is the damping constant of the actuator at joint. represents all the external forces and disturbance forces coupled from other joints motion. As for, it represents the muscle force computed from the neuromuscular-like control and applied to the actuator at joint. Assuming that joint has its initial joint position at, the muscle force generated by the neuromuscular-like controller can be rewritten from (2) to the following equation: (12) where represents the equilibrium position that muscle moves, and represents an input to the muscle after an LPF and can be represented as follows: LPF (13) In the above equation, is computed as (14) where is a positional motor command generated for joint from the planning, and is a reflex signal generated by the spindle-like model. In order to provide equilibrium position commands, is generated by the following equation: (15)

6 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 591 Fig. 4. Joint 1 involuntary movement (Km =781 N-m, H =2:065 rad/n-m). where is the planned desired joint position. As for the reflex signal, it is generated from the spindle-like model [(7) and (8)] at joint as follows: (16) where and are the internal position, velocity, and bias position of the spindle-like model at joint, is the reflex stiffness, and is a scaled damping coefficient for the spindle-like model. The properties of these parameters were discussed in the previous section. Biologically, the spindle-like model in the neuromuscularlike model emulates the function of spindle receptors [17]. This mechanism is capable of making automatic corrections in muscle toning to respond to external disturbance forces. It serves as a reflex sensor on the muscle force. When the muscle force rises past a threshold level, the spindle receptors will inhibit the motor unit, and turn the muscle activity off to prevent muscle from over-lengthening [26]. Meanwhile, it also regulates the muscle stiffness in voluntary movements. Based on this physiological phenomenon, the operation of the neuromuscular-like control has two modes, voluntary and involuntary, for free and constrained movements. In the voluntary mode for free movements, the motor commands defining the equilibrium joint positions can be computed from (15) based on the desired joint trajectory according to a planned path. If a voluntary movement is imposed by an external force, to adapt to the constraint, the transient equilibrium position will be shifted to comply the disturbance force. However, when the disturbance force dissipates, the equilibrium position will return to the desired target defined by motor commands. As for the involuntary mode, it is a pure force-following mode that has no positional motor commands from the planning. To emulate the effect of external force on muscle movement of this one-muscle model, the equivalent muscle movement,, in (12) and (16) needs to be remodeled as. After reversing the sign of the equivalent muscle movement, the reflex signal, generated by the spindle-like model in (16), will comply to the external disturbance force that changes. Then, the reflex signal will stimulate muscle mechanism and generate the muscle force at the joint actuator. As a result, the joint will move freely with the external force. Therefore, to have this involuntary model for force-following motion, an artificial Fig. 5. Joint 1 force enhancement in involuntary movement (Km = 781 N-m, H = 2:065 rad/n-m). control switch is needed for reversing the sign of equivalent muscle movement. By examining the neuromuscular-like control model carefully, it is noted that the spindle-reflex mechanism will generate the needed compliance for any motion, scaled through the reflex gain, in the muscle system. Without this nonlinear spindle-reflex mechanism, the muscle mechanism is simply a positional control mechanism. In summary, the internal spindle-reflex mechanism behaves as an impedance to comply to a changing displacement from the environment and then enhance the muscle force and provide the compliance for the motion. IV. EXPERIMENTS OF NEUROMUSCULAR-LIKE CONTROLLER ON PUMA 560 ROBOT In order to implement this neuromuscular-like control on each robot joint successfully, we will first simplify the number of changing parameters in the controller. Since the proposed spindle-reflex mechanism was experimentally modeled from the study of limb movements [17], we will assume that the spindle-reflex mechanism will be the same for all joints. In other words, the values of and in the model will be the same at all joints for any reflex motion. The parameter representing the length-tension effect will be kept to be one as in the original model to represent a linear feedback of joint position. The nonlinear effect of this length-tension gain will be explored in our future works. Particularly, the implementation of the neuromuscular-like control for each robot joint is simplified to adjust two parameters, the reflex gain and the muscle stiffness. The reflex gain will be adjusted for scaling different reflex forces represented by the discharge rate,. The muscle stiffness will be adjusted for different stiffnesses in needs of different joint muscles. In other words, these two parameters will be adjusted for controlling different free and constrained movements. A. Setup of PUMA 560 Robot System To implement the neuromuscular-like control on a PUMA 560 robot in our laboratory, we modified the original PUMA VAL controller. The modified system consists of a PC 486, a custom-made interface board, and a PUMA 560 arm. The custom-made interface board can switch the arm control from

7 592 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 Fig. 6. Joint 1 voluntary movement (Km = 781 N-m, H = 2:065 rad/n-m). Fig. 8. Joint 1 position errors of voluntary movement (Km = N-m, H =0:118 rad/n-m). Fig. 7. Joint 1 voluntary movement with disturbance force (Km = 781 N-m,H = 2:065 rad/n-m). the VAL controller to PC. The encoder information fed back from each PUMA joint is read through this interface board. The torque output computed from each neuromuscular-like joint controller implemented on PC is also sent through this board to each joint. The command planning strategies and robot forward and inverse kinematics of PUMA 560 robot are implemented in C language and running on our PC. At every sampling time, six neuromuscular-like joint controller will receive the motor command generated from our planning system and send joint muscle torques through the interface board to drive six PUMA joints. In order to drive individual joint, PUMA joints can also be locked through software control in PC so that any desired joint can be selected to move without interference from the other joints. Under this setup, the performance of neuromuscular-like controller will be examined with different gains for different free and constrained movements. B. Experiments on PUMA Joint 1 In this part of experiments, we try to identify a set of feasible gains for parameters in the neuromuscular-like control and their effectiveness to robotic control. To start our initial setup for gains, we converted the values of parameters in the neuromuscular-like model described in Section II to the rotating motion of PUMA robot. According to the kinematic and dynamic data of the robot and the dynamic data of the experiment setup for collecting movements [11], the following set of values are used in our model for starting up our initial experiments: rad/n-m, N-m/rad, N-m/rad, N-m/rad (s/rad) rad, and s. As discussed in the previous section, we will assume a fixed spindle-reflex mechanism for the controller and adjust the reflex gain and muscle-stiffness gain for generating various movements. To test the capability of a neuromuscular-like control on an industrial robot, we performed some experiments with joint 1 of PUMA 560 robot. The setup for testing involuntary movements on PUMA joint 1 was that we pulled or pushed joint 1 after all other five joints were locked. By pulling or pushing joint 1 alone, we could determine whether the joint torque is generated and enhanced by the neuromuscular-like control. If it is, the joint will follow the applied motion freely. The force-enhancement to observe is the property of the nonlinear damping effect multiplied by the short-range elasticity enhancement in the spindle-like model. This experiment tests the force-following capability of the neuromuscular-like controller without a force sensor because the spindle-like model will act as a reflex sensor. The servo sampling rate for the neuromuscular-like controller is about 1.5 ms measured by the PC486 clock. At each sampling time, the neuromuscularlike controller implemented on PC486 sends the computed torques through the interface board to PUMA controller that drives joint 1. After some experiments by adjusting the value of to rad/n-m and to 781 N-m/rad, the response of PUMA s joint 1 was amazingly good and smooth. The result in Fig. 4 shows the involuntary movement of joint 1 followed by an external force applied to the arm. Fig. 5 depicts the enhanced muscle forces versus the positional changes of joint 1. Since this experiment had no input motor commands, the internal spindle-reflex mechanism will comply to the changes in joint position and generated the muscle force to provide the compliance for the involuntary motion. As demonstrated in Fig. 5, the muscle force was generated and enhanced by the spindle-like model to move the joint to follow the external force. The force enhancement with a big jump can be observed at the beginning of the movement (from 90 to 87 ). Conversely, when the external force diminished gradually from

8 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 593 (a) (b) (c) Fig. 9. PUMA 560 is commanded to insert a peg into the outlet tube of a funnel. (a) Projection of tip position on X-Z plane. (b) Projection of tip position on Y -Z plane. (c) The trajectory of tip in Cartesian space. 60 to 29, the muscle force should diminish accordingly. Therefore, the limb motion should stop rapidly (reversibly enhanced) at the end of the motion as demonstrated in the figure. Due to this force enhancement property provided by the spindle-like model, no extra force sensor for designing a force control is required for the neuromuscular-like control to have force-following capability. And more significantly, it demonstrated that a compliance control can be designed for a robot without using any force sensor. To generate voluntary movements, we used the same gain set as in testing the involuntary movement. The desired motor commands for the intermediate joint positions at joint 1 were generated by (9) based on the planning strategy of minimumjerk. The experimental result of moving 100 is plotted in Fig. 6. In this case the maximum position error is about 2 in the middle of the movement. The error at the end is about 0.9. To test the compliant capability of the control, we did another experiment. With the same voluntary movement, during the joint motion, we could stop the joint with our hand, easily. Then, after we released the joint for holding it about 3 s, an amazing result was that the joint moved to its destination without any problem. In other words, the developed controller attempts to make up the lost time by following the original trajectory. This experiment further demonstrated that a robot joint controlled by the neuromuscular-like model can be compliant and adapt to its environment. To further improve the accuracy of voluntary movements, we also performed another experiment by increasing the gain to N-m/rad and decreasing the gain of to rad/n-m. The position errors along the same movement are plotted in Fig. 8. The maximum positional error is less than 0.1 and the settling error in position is less than However, for such a strong stiffness, the joint motion could not be stopped during the movement as we demonstrated in the previous experiment. This experiment showed that a precise positional control for free voluntary movements can also be designed with the neuromuscular-like controller in addition to the compliant capability of force-following. In summary, our experiments with joint 1 of PUMA 560 robot demonstrated that the neuromuscular-like controller can have different stiffness for free voluntary movements, constrained involuntary movements and compliant movements (partially constrained) by adjusting the reflex gain and the muscle-stiffness gain. The product of these two parameters within the control loop provides certain degree of force enhancement for muscle force. With the same degree of force enhancement, the higher the reflex gain is, the more compliant the system will be. On the other hand, the higher the stiffness

9 594 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 (a) (b) (c) Fig. 10. Robot arm is led by hand to trace a circle drawn on a table. (a) Projection of tip position on X-Y plane. (b) Projection of tip position on Y -Z plane. (c) The trajectory of tip in Cartesian space. gain is, the stiffer the system will be. These experimental results strongly proved that the neuromuscular-like control has the advantage in providing compliance for the constrained tasks. In addition, the spindle-like model in the neuromuscularlike control will generate the needed compliant force from changes in position. This reflex force will then generate the muscle force through the product of reflex gain and muscle stiffness gain. As a result, an expensive force sensor is not required for providing the force feedback in the system. This is a great cost saving in designing a compliance control. In the next section, we will explore this property further for improving robotic control in compliance. C. Experiments of Compliance on Six Joints of PUMA 560 Robot During our experiments with PUMA joint 1, we found that different parameters of the neuromuscular-like controller control different functions. The reflex gain will affect the capability of the system in adaptation. The muscle stiffness gain will affect the errors in position tracking. These two parameters have opposite effects on the system. One will increase the compliant capability of the system and the other will increase the stiffness of the system for a better position control. As for the spindle-like mechanism, it simply performs an amazing task of converting changing positions to compliant forces. Because of its capability, no force sensor is needed in the system. To set up the parameters of the neuromuscular-like controller for all joints, the same approach and experiments as we did for joint 1 were also applied to the other five joints of PUMA robot. To have all six neuromuscular-like controllers running on the PC486 for a Cartesian task, the sampling time for sending joint torques requires 2.4 ms to finish all computations needed for planning, gravity compensation, control, and data transfer. Although the 2.4-ms servo sampling time is longer than we desired for controlling our robot, the performance is acceptable. The servo sampling time can be tremendously shortened if a Pentium PC is used. By adjusting the reflex and muscle-stiffness gains for all six joint controllers, the following three experiments with different constrained tasks were performed by the PUMA robot. To demonstrate the compliance control for a constrained motion, three plots in Fig. 9 show the results of an experiment that PUMA robot inserted a peg into the outlet tube of a funnel and the neuromuscular-like controllers adjusted the tip path for insertion. The result demonstrated that robot s

10 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 595 (a) (b) (c) (d) (e) (f) Fig. 11. Robot is pulled away by a disturbance force while it is approaching to the target point. (a) Joint 1 trajectory. (b) Joint 2 trajectory. (c) Joint 3 trajectory. (d) Joint 4 trajectory. (e) Joint 5 trajectory. (f) Joint 6 trajectory. tip was compliant to the constraint caused by the contacting surface. To prove the capability of a pure force-following for constrained tasks, three plots in Fig. 10 show the results of an experiment that robot was hand-led to roughly trace a circular object. To further verify the compliant capability of the controller, six plots in Fig. 11 illustrate the responses of six joints of the PUMA 560 robot when the robot adapted to a stopping force during a voluntary movement and then reached its target position after the constrained force was removed. This experiment further substantiated that a robot controlled by six neuromuscular-like joint controllers can comply and adapt to its environment. V. CONCLUSION Limitations in current robot compliance controls motivates the study of the neuromuscular system and limb movements. Based on the properties and mechanisms of the musclereflex system, a neuromuscular-like model was formulated

11 596 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 5, NO. 6, NOVEMBER 1997 for control. A very important property of the developed neuromuscular-like control is that it provides the needed compliance for various movements. The capability in performing free and constrained movements demonstrated that a neuromuscular-like control is very useful for robotic applications requiring adaptation. To confirm our findings from the biological muscle-reflex system, we use a PUMA 560 robot to demonstrate the proposed controller. The experimental results showed that the same neuromuscular-like controller can generate free voluntary movements, constrained involuntary movements, and partially constrained compliant movements by adjusting the values of two parameters, the reflex gain and the musclestiffness gain. The product of these two parameters within the control loop provides certain degree of force enhancement for muscle force. With the same degree of force enhancement, the higher the reflex gain is, the more compliant the system will be. On the other hand, the higher the stiffness gain is, the stiffer the system will be. These experimental results strongly proved that the proposed controller has the advantage in providing compliance for the constrained tasks. Another important property of the neuromuscular-like control is that the emulated spindle-reflex mechanism will comply to the sensed changes in position. This reflex force will then generate the muscle force through the product of reflex gain and muscle stiffness gain. As a result, no expensive force sensor is required for providing the force feedback in the system. This is a great cost saving in designing a compliance control. Due to the capability of current industrial robots, applications of robots have been limited. Although many approaches were proposed to solve the problem, the mechanism of primate limb seems to be the best solution. Therefore, we believe that the success of this research in linking the findings in the biological muscle system to current robotic control will advance robotic applications to the next level requiring adaptation. [10] C. C. A. M. Gielen, J. C. Houk, S. L. Marcus, and L. E. Miller, Viscoelastic properties of the wrist motor servo in man, Ann. Biomed. Eng., vol. 12, pp , [11] L. E. Miller, Reflex stiffness of the human wrist, M.S. thesis, Dept. Physiology, Northwestern Univ., Evanston, IL, [12] C. H. Wu, J. C. Houk, K. Y. Young, and L. E. Miller, Nonlinear damping of limb motion, in Multiple Muscle Systems: Biomechanics and Movement Organization, J. M. Winters and S. L.-Y. Woo, Eds. New York: Springer-Verlag, 1990, pp [13] G. I. Zahalak, A Distribution Moment approximation for kinetic theories of muscular contraction, Math. Biosci., vol. 55, pp , [14] A. F. Huxley, Muscle structure and theories of constraction, Prog. Biophys. Chem., vol. 7, pp , [15] C. C. A. M. Gielen and J. C. Houk, A Model of the motor servo: Incorporating nonlinear spindle receptor and muscle mechanical properties, Biol. Cybern., vol. 57, pp , [16] Z. Hasan, A model of spindle afferent response to muscle stretch, J. Neurophysiology, vol. 49, pp , [17] C. H. Wu, K. Y. Young, and J. C. Houk, A neuromuscular-like model for robotic compliance control, in Proc IEEE Int. Conf. Robot. Automat., Cincinnati, OH, 1990, pp [18] C. H. Wu, K. Y. Young, and K. S. Hwang, Analysis of voluntary movements for robotic control, in Proc. IEEE 1991 Int. Conf. Robot. Automat., Sacramento, CA, 1991, pp [19] C. H. Wu, K. Y. Young, K. S. Hwang, and S. Lehman, Analysis of voluntary movements for robotic control, IEEE Contr. Syst. Mag., vol. 2, pp. 8 14, [20] E. Bizzi, N. Accornero, W. Chapple, and N. Hogan, Posture control and trajectory formation during arm movement, J. Neurosci., vol. 4, pp , [21] T. Flash and N. Hogan, The coordination of arm movements: An experimentally confirmed mathematical model, J. Neurosci., vol. 5, pp , [22] J. M. Hollerbach and C. G. Atkeson, Deducing planning variables from experimental arm trajectories: Pitfalls and possibilities, Biol. Cybern., vol. 56, pp , [23] T. Flash, The control of hand equilibrium trajectories in multijoint arm movements, Biol. Cybern., vol. 57, pp , [24] N. Hogan, The mechanics of multijoint posture and movement control, Biol. Cybern., vol. 52, pp , [25] Y. Uno, M. Kawato, and R. Suzuki, Formation and control of optimal trajectory in human multijoint arm movement, Biol. Cybern., vol. 61, pp , [26] A. T. McMahon, Muscles, Reflexes, and Locomotion. Princeton, NJ: Princeton Univ. Press, REFERENCES [1] C. H. Wu, Compliance control of robot manipulators based on joint torque servo, Int. J. Robot. Res., vol. 4, pp , [2] C. H. Wu, Compliance, in International Encyclopedia of Robotics: Application and Automation, vol. 1. New York: Wiley, 1988, pp [3] P. R. Paul and B. Simano, Compliance and control, in Proc Joint Automat. Contr. Conf., San Francisco, CA, 1976, pp [4] M. H. Raibert and J. J. Craig, Hybrid position/force control of manipulator, J. Dynamic Syst., Measurement, Contr., vol. 102, pp , [5] D. E. Whitney, Force feedback control of manipulator fine motion, Trans. ASME, J. Dynamic Syst., Measurement, Contr., pp , [6] J. K. Salisbury, Active stiffness control of a manipulator in Cartesian coordinates, Proc. IEEE Conf. Decision Contr., Albuquerque, NM, 1980, pp [7] N. Hogan, Impedance control: An approach to manipulation Part I: Theory; Part II: Implementation; Part III: Application, J. Dynamic Syst., Measurement, Contr., vol. 107, pp. 1 24, [8] S. C. Jacobsen, H. Ko, E. K. Iversen, and C. C. Davis, Control strategies for tendon-driven manipulators, IEEE Contr. Syst. Mag., vol. 10, pp , [9] J. C. Houk and W. Z. Rymer, Neural control of muscle length and tension, in Handbook of Physiology The Nervous System II, vol. 2. Bethesda, MD: Amer. Physiological Soc., 1981, pp Chi-haur Wu (S 79 M 80 SM 91) received the B.S. degree in electrical engineering from National Taiwan University, Taiwan, in 1973, the M.S. degree in electrical engineering from Virginia Polytechnic Institute and State University, Blacksburg, in 1977, and the Ph.D. degree in electrical engineering from Purdue University, West Lafayette, IN, in He joined Unimation Inc., Danbury, CT, as a Senior Engineer. His job involved in designing robot motion control algorithms and digital servo systems for PUMA robots and hydraulic-servo Unimate robots. Since September 1983, he has been an Associate Professor of Electrical and Computer Engineering at Northwestern University, Evanston, IL. His areas of interest are robotics, limb control and compliance control, rehabilitation robot, part assembly strategy, computer graphics and CAD/CAM industrial automation, computer-assisted surgical robot system, medical instrumentation and automation, layered manufacturing, control of smart structures, vibration and noise control, and active suspension control. During 1985, Dr. Wu was honored with the Outstanding Young Manufacturing Engineer Award from the Society of Manufacturing Engineers. He was a member of the Organized Committee of the 1992 IEEE International Conference on Systems, Man, and Cybernetics and the 1993 IEEE International Symposium on Intelligent Control. He was also a member of the Program Committee of the 1993 IEEE International Conference on Robotics and Automation, the 1995 IEEE International Conference on Decision and Control and the 1996 IEEE International Conference on Robotics and Automation.

12 WU et al.: NEUROMUSCULAR-LIKE CONTROL FOR ROBOTIC COMPLIANCE 597 Kao-Shing Hwang (S 92 M 92) received the B.S. Degree in Industrial Design from National Cheng Kung University, Taiwan, in 1981, and the M.M.E. and Ph.D. degrees in electrical engineering and computer science from Northwestern University, Evanston, I.L., in 1989 and 1993, respectively. He was a Design Engineer at the SANYO Electric Co., Taipei, Taiwan during Between 1987 and 1988, he joined the C & D Microsystem Co., Plastow, NH, as a System Programmer. His job involved designing PC drivers, and graphic animation. Since August 1993, he has been with National Chung Cheng University, Taiwan, where he is an Associate Professor. He is also the Director of the Information Management Division at the university computer center. His interests include neural networks and learning control, robotic compliance, and collision avoidance. Shih-Lang Chang received the B.S. degree in mechanical engineering from National Chiao Tung University, Hsinchu, Taiwan, in 1989, and the M.S. degree from Northwestern University, Evanston, IL, in He is currently a Ph.D. candidate of mechanical engineering in Northwestern University. During , he was an Ordinance Lieutenant in the Taiwan military while performing his obligatory service. Since 1995, he has been a Research Assistant in the Rehabilitation Institute of Chicago, where he is engaged in research of human limb dynamics. His research interests include learning control, flexible manufacturing systems, and compliance control. Mr. Chang is a member of ASME and SAE.

Estimation of the Upper Limb Lifting Movement Under Varying Weight and Movement Speed

Estimation of the Upper Limb Lifting Movement Under Varying Weight and Movement Speed 1 Sungyoon Lee, 1 Jaesung Oh, 1 Youngwon Kim, 1 Minsuk Kwon * Jaehyo Kim 1 Department of mechanical & control engineering, Handong University, qlfhlxhl@nate.com * Department of mechanical & control engineering,

More information

Multi-joint Mechanics Dr. Ted Milner (KIN 416)

Multi-joint Mechanics Dr. Ted Milner (KIN 416) Multi-joint Mechanics Dr. Ted Milner (KIN 416) Muscle Function and Activation It is not a straightforward matter to predict the activation pattern of a set of muscles when these muscles act on multiple

More information

EMG-Driven Human Model for Orthosis Control

EMG-Driven Human Model for Orthosis Control EMG-Driven Human Model for Orthosis Control Christian Fleischer, Günter Hommel Institute for Computer Engineering and Microelectronics Berlin University of Technology, Germany {fleischer, hommel}@cs.tu-berlin.de

More information

Q: What is the relationship between muscle forces and EMG data that we have collected?

Q: What is the relationship between muscle forces and EMG data that we have collected? FAQs ABOUT OPENSIM Q: What is the relationship between muscle forces and EMG data that we have collected? A: Muscle models in OpenSim generate force based on three parameters: activation, muscle fiber

More information

Kinematic and Dynamic Adaptation of

Kinematic and Dynamic Adaptation of Humanoids 2008 Workshop on Imitation and Coaching in Humanoid Robots Kinematic and Dynamic Adaptation of Human Motion for Imitation Katsu Yamane Disney Research, Pittsburgh Carnegie Mellon University Programming

More information

Study on the control of variable resistance for isokinetic muscle training system

Study on the control of variable resistance for isokinetic muscle training system Technology and Health Care 25 (2017) S45 S52 DOI 10.3233/THC-171305 IOS Press S45 Study on the control of variable resistance for isokinetic muscle training system Lan Wang, Zhenyuan Zhang, Yi Yu and Guangwei

More information

ANTICIPATING DYNAMIC LOADS IN HANDLING OBJECTS.

ANTICIPATING DYNAMIC LOADS IN HANDLING OBJECTS. ANTICIPATING DYNAMIC LOADS IN HANDLING OBJECTS. Alan M. Wing Sensory Motor Neuroscience Centre, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK. a.m.wing@bham.ac.uk J. Randall Flanagan

More information

Re-establishing establishing Neuromuscular

Re-establishing establishing Neuromuscular Re-establishing establishing Neuromuscular Control Why is NMC Critical? What is NMC? Physiology of Mechanoreceptors Elements of NMC Lower-Extremity Techniques Upper-Extremity Techniques Readings Chapter

More information

Fast Simulation of Arm Dynamics for Real-time, Userin-the-loop. Ed Chadwick Keele University Staffordshire, UK.

Fast Simulation of Arm Dynamics for Real-time, Userin-the-loop. Ed Chadwick Keele University Staffordshire, UK. Fast Simulation of Arm Dynamics for Real-time, Userin-the-loop Control Applications Ed Chadwick Keele University Staffordshire, UK. Acknowledgements Dimitra Blana, Keele University, Staffordshire, UK.

More information

Five-Fingered Assistive Hand with Mechanical Compliance of Human Finger

Five-Fingered Assistive Hand with Mechanical Compliance of Human Finger 28 IEEE International Conference on Robotics and Automation Pasadena, CA, USA, May 19-23, 28 Five-Fingered Assistive Hand with Mechanical Compliance of Human Finger Yasuhisa Hasegawa, Yasuyuki Mikami,

More information

Human Arm Posture Control Using the Impedance Controllability of the Musculo-Skeletal System Against the Alteration of the Environments

Human Arm Posture Control Using the Impedance Controllability of the Musculo-Skeletal System Against the Alteration of the Environments Transactions on Control, Automation, and Systems Engineering Vol. 4, No. 1, March, 00 43 Human Arm Posture Control Using the Impedance Controllability of the Musculo-Skeletal System Against the Alteration

More information

Anatol G. Feldman. Referent control of action and perception. Challenging conventional theories in behavioral neuroscience

Anatol G. Feldman. Referent control of action and perception. Challenging conventional theories in behavioral neuroscience Anatol G. Feldman Referent control of action and perception Challenging conventional theories in behavioral neuroscience Referent control of action and perception Anatol G. Feldman Referent control of

More information

Biomechanics (part 2)

Biomechanics (part 2) Biomechanics (part 2) MCE 493/593 & ECE 492/592 Prosthesis Design and Control September 11, 214 Antonie J. (Ton) van den Bogert Mechanical Engineering Cleveland State University 1 Today Coupling between

More information

The Physiology of the Senses Chapter 8 - Muscle Sense

The Physiology of the Senses Chapter 8 - Muscle Sense The Physiology of the Senses Chapter 8 - Muscle Sense www.tutis.ca/senses/ Contents Objectives... 1 Introduction... 2 Muscle Spindles and Golgi Tendon Organs... 3 Gamma Drive... 5 Three Spinal Reflexes...

More information

NZQA Expiring unit standard 7026 version 4 Page 1 of 7. Apply knowledge of functional anatomy and biomechanics

NZQA Expiring unit standard 7026 version 4 Page 1 of 7. Apply knowledge of functional anatomy and biomechanics Page 1 of 7 Title Apply knowledge of functional anatomy and biomechanics Level 5 Credits 5 Purpose People credited with this unit standard are able to: apply knowledge of human anatomy relevant to exercise

More information

A computational model for optimal muscle activity considering muscle viscoelasticity in wrist movements

A computational model for optimal muscle activity considering muscle viscoelasticity in wrist movements J Neurophysiol 19: 2145 216, 213. First published January 16, 213; doi:1.1152/jn.542.211. A computational model for optimal muscle activity considering muscle viscoelasticity in wrist movements Hiroyuki

More information

学位規則第 9 条第 2 項により要約公開 ; 許諾条件により要約 Rightは に公開 ; 許諾条件により要旨は 公開

学位規則第 9 条第 2 項により要約公開 ; 許諾条件により要約 Rightは に公開 ; 許諾条件により要旨は 公開 Title Muscle synergy for coordinating red Digest_ 要約 ) Author(s) Hagio, Shota Citation Kyoto University ( 京都大学 ) Issue Date 2016-03-23 URL https://doi.org/10.14989/doctor.k19 学位規則第 9 条第 2 項により要約公開 ; 許諾条件により要約

More information

Control principles in upper-limb prostheses

Control principles in upper-limb prostheses Control principles in upper-limb prostheses electromyographic (EMG) signals generated by muscle contractions electroneurographic (ENG) signals interface with the peripheral nervous system (PNS) interface

More information

298 SECTION V: THE HAND

298 SECTION V: THE HAND Section V THE HAND INTRODUCTION As a young neurologist, I was curious about the various neurological signs of motor discoordination evidenced by human patients with Parkinson's disease, cerebellar disease,

More information

Learning Classifier Systems (LCS/XCSF)

Learning Classifier Systems (LCS/XCSF) Context-Dependent Predictions and Cognitive Arm Control with XCSF Learning Classifier Systems (LCS/XCSF) Laurentius Florentin Gruber Seminar aus Künstlicher Intelligenz WS 2015/16 Professor Johannes Fürnkranz

More information

Design and Dynamic Modeling of Flexible Rehabilitation Mechanical Glove

Design and Dynamic Modeling of Flexible Rehabilitation Mechanical Glove IOP Conference Series: Materials Science and Engineering PAPER OPEN ACCESS Design and Dynamic Modeling of Flexible Rehabilitation Mechanical Glove To cite this article: M X Lin et al 2018 IOP Conf. Ser.:

More information

SINGLE- AND TWO-JOINT MOVEMENTS IN HUMANS

SINGLE- AND TWO-JOINT MOVEMENTS IN HUMANS SINGLE- AND TWO-JOINT MOVEMENTS IN HUMANS BASIC PRINCIPLES OF THE MOVEMENT ANALYSIS METHODS OF THE MOVEMENT ANALYSIS EMGs are recorded by pairs of the surface electrodes with center to center distance

More information

Kota Samarahan, Sarawak, Malaysia. Keywords: inverse dynamics analysis, musculoskeletal model, muscle activity, intradiscal force

Kota Samarahan, Sarawak, Malaysia. Keywords: inverse dynamics analysis, musculoskeletal model, muscle activity, intradiscal force Advanced Engineering Forum Online: 213-12-3 ISSN: 2234-991X, Vol. 1, pp 373-378 doi:1.428/www.scientific.net/aef.1.373 213 Trans Tech Publications, Switzerland Musculoskeletal analysis of driving fatigue:

More information

HUMAN MOTOR CONTROL. Emmanuel Guigon

HUMAN MOTOR CONTROL. Emmanuel Guigon HUMAN MOTOR CONTROL Emmanuel Guigon Institut des Systèmes Intelligents et de Robotique Université Pierre et Marie Curie CNRS / UMR 7222 Paris, France emmanuel.guigon@upmc.fr e.guigon.free.fr/teaching.html

More information

A Cooperatively Controlled Robot for Ultrasound Monitoring of Radiation Therapy

A Cooperatively Controlled Robot for Ultrasound Monitoring of Radiation Therapy A Cooperatively Controlled Robot for Ultrasound Monitoring of Radiation Therapy H. Tutkun Şen 1, Muyinatu A. Lediju Bell 1, Iulian Iordachita 2, John Wong 3 and Peter Kazanzides 1 Abstract Image-guided

More information

Vibration Analysis of Finger Using Non linear FEM to Understand HAV Syndrome

Vibration Analysis of Finger Using Non linear FEM to Understand HAV Syndrome Vibration Analysis of Finger Using Non linear FEM to Understand HAV Syndrome Shrikant Pattnaik, Jay Kim Dept. of Mechanical Engineering University of Cincinnati What is HAVS Hand Arm Vibration Syndrome

More information

OPEN QUESTIONS IN COMPUTATIONAL MOTOR CONTROL

OPEN QUESTIONS IN COMPUTATIONAL MOTOR CONTROL September 13, 2011 10:05:27am WSPC/179-JIN 00274 ISSN: 0219-6352 FA2 Journal of Integrative Neuroscience, Vol. 10, No. 3 (2011) 391 417 c Imperial College Press DOI: 10.1142/S0219635211002749 OPEN QUESTIONS

More information

2282. Design of a bionic-inspired exoskeleton robot for lower limb assist

2282. Design of a bionic-inspired exoskeleton robot for lower limb assist 2282. Design of a bionic-inspired exoskeleton robot for lower limb assist Yun-Ping Sun 1, Shuo-Ching Chen 2, Yen-Chu Liang 3, Lung-Nan Wu 4 1 Department of Mechanical Engineering, Cheng Shiu University,

More information

Comparison of Robot-Assisted Reaching to Free Reaching in Promoting Recovery From Chronic Stroke

Comparison of Robot-Assisted Reaching to Free Reaching in Promoting Recovery From Chronic Stroke Comparison of Robot-Assisted Reaching to Free Reaching in Promoting Recovery From Chronic Stroke Leonard E. Kahn, M.S. 1,2, Michele Averbuch, P.T. 1, W. Zev Rymer, M.D., Ph.D. 1,2, David J. Reinkensmeyer,

More information

Chapter 13. The Nature of Muscle Spindles, Somatic Reflexes, and Posture

Chapter 13. The Nature of Muscle Spindles, Somatic Reflexes, and Posture Chapter 13 The Nature of Muscle Spindles, Somatic Reflexes, and Posture Nature of Reflexes A reflex is an involuntary responses initiated by a sensory input resulting in a change in the effecter tissue

More information

Hand of Hope. For hand rehabilitation. Member of Vincent Medical Holdings Limited

Hand of Hope. For hand rehabilitation. Member of Vincent Medical Holdings Limited Hand of Hope For hand rehabilitation Member of Vincent Medical Holdings Limited Over 17 Million people worldwide suffer a stroke each year A stroke is the largest cause of a disability with half of all

More information

REACTION TIME MEASUREMENT APPLIED TO MULTIMODAL HUMAN CONTROL MODELING

REACTION TIME MEASUREMENT APPLIED TO MULTIMODAL HUMAN CONTROL MODELING XIX IMEKO World Congress Fundamental and Applied Metrology September 6 11, 2009, Lisbon, Portugal REACTION TIME MEASUREMENT APPLIED TO MULTIMODAL HUMAN CONTROL MODELING Edwardo Arata Y. Murakami 1 1 Digital

More information

Biomechanics of Skeletal Muscle

Biomechanics of Skeletal Muscle Biomechanics of Skeletal Muscle Contents I. Composition & structure of skeletal muscle II. Mechanics of Muscle Contraction III. Force production in muscle IV. Muscle remodeling V. Summary 2 Muscle types:

More information

Dynamic Rule-based Agent

Dynamic Rule-based Agent International Journal of Engineering Research and Technology. ISSN 0974-3154 Volume 11, Number 4 (2018), pp. 605-613 International Research Publication House http://www.irphouse.com Dynamic Rule-based

More information

Excitation-Contraction Coupling & Reflexes, Proprioception and Movement. PSK 4U Unit 4, Day 4

Excitation-Contraction Coupling & Reflexes, Proprioception and Movement. PSK 4U Unit 4, Day 4 Excitation-Contraction Coupling & Reflexes, Proprioception and Movement PSK 4U Unit 4, Day 4 Excitation-Contraction Coupling Muscles work by converting electrical and chemical energy into mechanical energy!

More information

KINEMATIC ANALYSIS OF ELBOW REHABILITATION EQUIPMENT

KINEMATIC ANALYSIS OF ELBOW REHABILITATION EQUIPMENT Bulletin of the Transilvania University of Braşov Vol. 10 (59) No. 2-2017 Series I: Engineering Sciences KINEMATIC ANALYSIS OF ELBOW REHABILITATION EQUIPMENT G. VETRICE 1 A. DEACONESCU 1 Abstract: The

More information

Muscle-Tendon Mechanics Dr. Ted Milner (KIN 416)

Muscle-Tendon Mechanics Dr. Ted Milner (KIN 416) Muscle-Tendon Mechanics Dr. Ted Milner (KIN 416) Muscle Fiber Geometry Muscle fibers are linked together by collagenous connective tissue. Endomysium surrounds individual fibers, perimysium collects bundles

More information

IDENTIFYING the mechanical properties of the human

IDENTIFYING the mechanical properties of the human IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 46, NO. 4, APRIL 1999 409 A Robust Ensemble Data Method for Identification of Human Joint Mechanical Properties During Movement Yangming Xu, Member, IEEE,

More information

Optimisation of high bar circling technique for consistent performance of a triple piked somersault dismount

Optimisation of high bar circling technique for consistent performance of a triple piked somersault dismount Loughborough University Institutional Repository Optimisation of high bar circling technique for consistent performance of a triple piked somersault dismount This item was submitted to Loughborough University's

More information

IEEE/ASME TRANSACTIONS ON MECHATRONICS, VOL. 2, NO. 4, DECEMBER

IEEE/ASME TRANSACTIONS ON MECHATRONICS, VOL. 2, NO. 4, DECEMBER IEEE/ASME TRANSACTIONS ON MECHATRONICS, VOL. 2, NO. 4, DECEMBER 1997 237 Control of Smart Exercise Machines Part I: Problem Formulation and Nonadaptive Control Perry Y. Li, Member, IEEE, and Roberto Horowitz,

More information

The Biomechanics of Human Skeletal Muscle

The Biomechanics of Human Skeletal Muscle AML2506 Biomechanics and Flow Simulation Day 03B The Biomechanics of Human Skeletal Muscle Session Speaker Dr. M. D. Deshpande 1 Session Objectives At the end of this session the delegate would have understood

More information

STIFFNESS OF THE HUMAN LIPS IN PARKINSON S DISEASE

STIFFNESS OF THE HUMAN LIPS IN PARKINSON S DISEASE Lana Seibel, Steven Barlow, Michael Hammer, Shiva Prasad, & Rajesh Pahwa 1 Communication Neuroscience Laboratories Department of Speech-Language-Hearing: Sciences and Disorders 3001 Dole Human Development

More information

Modeling the muscular response to motor neuron spike-trains. Laura Miller and Katie Newhall SAMSI Transition Workshop May 4, 2016

Modeling the muscular response to motor neuron spike-trains. Laura Miller and Katie Newhall SAMSI Transition Workshop May 4, 2016 Modeling the muscular response to motor neuron spike-trains Laura Miller and Katie Newhall SAMSI Transition Workshop May 4, 2016 Outline 1. Motivation for an integrative neural and mechanical view of animal

More information

Robot control using electromyography (EMG) signals of the wrist

Robot control using electromyography (EMG) signals of the wrist Robot control using electromyography (EMG) signals of the wrist C. DaSalla,J.Kim and Y. Koike,2 Tokyo Institute of Technology, R2-5, 4259 Nagatsuta-cho, Midori-ku, Yokohama 226-853, Japan 2 CREST, Japan

More information

Navigator: 2 Degree of Freedom Robotics Hand Rehabilitation Device

Navigator: 2 Degree of Freedom Robotics Hand Rehabilitation Device Navigator: 2 Degree of Freedom Robotics Hand Rehabilitation Device Design Team Ray Adler, Katherine Bausemer, Joseph Gonsalves, Patrick Murphy, Kevin Thompson Design Advisors Prof. Constantinos Mavroidis,

More information

SITES OF FAILURE IN MUSCLE FATIGUE

SITES OF FAILURE IN MUSCLE FATIGUE of 4 SITES OF FAILURE IN MUSCLE FATIGUE Li-Qun Zhang -4 and William Z. Rymer,2,4 Sensory Motor Performance Program, Rehabilitation Institute of Chicago Departments of 2 Physical Medicine and Rehabilitation,

More information

Musculoskeletal System. Terms. Origin (Proximal Attachment) Insertion (Distal Attachment)

Musculoskeletal System. Terms. Origin (Proximal Attachment) Insertion (Distal Attachment) Musculoskeletal System Terms Origin (Proximal Attachment) Insertion (Distal Attachment) Agonist- prime mover Antagonist- provides a braking force Synergist- assists indirectly in the movement Musculoskeletal

More information

Thesis Rehabilitation robotics (RIA) Robotics for Bioengineering Forefront research at PRISMA Lab and ICAROS Center

Thesis Rehabilitation robotics (RIA) Robotics for Bioengineering Forefront research at PRISMA Lab and ICAROS Center Thesis Rehabilitation robotics (RIA) RIA-1. Mechanical design of sensorized and under-actuated artificial hands with simulation and/or prototype tests The thesis work involves the study of kinematics of

More information

Basics of kinetics. Kinesiology RHS 341 Lecture 7 Dr. Einas Al-Eisa

Basics of kinetics. Kinesiology RHS 341 Lecture 7 Dr. Einas Al-Eisa Basics of kinetics Kinesiology RHS 341 Lecture 7 Dr. Einas Al-Eisa Mass The amount of matter in an object Weight A force, which depends on the mass and acceleration Free-body analysis A technique of looking

More information

Goals. Musculoskeletal Model. Human Body Modeling. Muscles/Tendons/Ligaments. Skeleton 7/23/2013

Goals. Musculoskeletal Model. Human Body Modeling. Muscles/Tendons/Ligaments. Skeleton 7/23/2013 Goals Robotics and Animatronics in Disney Lecture 7: Human Modeling and Control Katsu Yamane kyamane@disneyresearch.com Introduce basic physiology and describe how to model it using robotics techniques

More information

CONTROL OF THE BOUNDARY CONDITIONS OF A DYNAMIC KNEE SIMULATOR

CONTROL OF THE BOUNDARY CONDITIONS OF A DYNAMIC KNEE SIMULATOR CONTROL OF THE BOUNDARY CONDITIONS OF A DYNAMIC KNEE SIMULATOR J. Tiré 1, J. Victor 2, P. De Baets 3 and M.A. Verstraete 2 1 Ghent University, Belgium 2 Ghent University, Department of Physical Medicine

More information

Lesson 6.4 REFLEXES AND PROPRIOCEPTION

Lesson 6.4 REFLEXES AND PROPRIOCEPTION Lesson 6.4 REFLEXES AND PROPRIOCEPTION (a) The Reflex Arc ~ ~ ~ TOPICS COVERED IN THIS LESSON (b) Proprioception and Proprioceptors 2015 Thompson Educational Publishing, Inc. 1 What Are Reflexes? Reflexes

More information

Chapter 6. Results. 6.1 Introduction

Chapter 6. Results. 6.1 Introduction Chapter 6 Results 6.1 Introduction This chapter presents results of both optimization and characterization approaches. In the optimization case, we report results of an experimental study done with persons.

More information

Making Things Happen: Simple Motor Control

Making Things Happen: Simple Motor Control Making Things Happen: Simple Motor Control How Your Brain Works - Week 10 Prof. Jan Schnupp wschnupp@cityu.edu.hk HowYourBrainWorks.net The Story So Far In the first few lectures we introduced you to some

More information

OpenSim Tutorial #2 Simulation and Analysis of a Tendon Transfer Surgery

OpenSim Tutorial #2 Simulation and Analysis of a Tendon Transfer Surgery OpenSim Tutorial #2 Simulation and Analysis of a Tendon Transfer Surgery Laboratory Developers: Scott Delp, Wendy Murray, Samuel Hamner Neuromuscular Biomechanics Laboratory Stanford University I. OBJECTIVES

More information

REHABILITATION program is the main stay of treatment

REHABILITATION program is the main stay of treatment IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, VOL. 13, NO. 3, SEPTEMBER 2005 349 A Rehabilitation Robot With Force-Position Hybrid Fuzzy Controller: Hybrid Fuzzy Control of Rehabilitation

More information

Information Processing During Transient Responses in the Crayfish Visual System

Information Processing During Transient Responses in the Crayfish Visual System Information Processing During Transient Responses in the Crayfish Visual System Christopher J. Rozell, Don. H. Johnson and Raymon M. Glantz Department of Electrical & Computer Engineering Department of

More information

Motor Control in Biomechanics In Honor of Prof. T. Kiryu s retirement from rich academic career at Niigata University

Motor Control in Biomechanics In Honor of Prof. T. Kiryu s retirement from rich academic career at Niigata University ASIAN SYMPOSIUM ON Motor Control in Biomechanics In Honor of Prof. T. Kiryu s retirement from rich academic career at Niigata University APRIL 20, 2018 TOKYO INSTITUTE OF TECHNOLOGY Invited Speakers Dr.

More information

International Journal on Bioinformatics & Biosciences (IJBB) Vol.6, No.3/4, December 2016

International Journal on Bioinformatics & Biosciences (IJBB) Vol.6, No.3/4, December 2016 EFFECT OF POSTURAL CONTROL BIOMECHANICAL GAIN ON PSYCHOPHYSICAL DETECTION THRESHOLDS IN ANTERIOR HORIZONTAL TRANSLATION OF STANDING BLINDFOLDED SUBJECTS Shahrokh N Sani 1 and Charles J Robinson 2 1 Department

More information

Smart. Training. Developing advanced exercise machines

Smart. Training. Developing advanced exercise machines PAGE 24 CUSTOMERS Developing advanced exercise machines Smart Training Researchers from Cleveland State University are developing new kinds of exercise machines for athletic conditioning, rehabilitation

More information

NEUROLOGICAL injury is a leading cause of permanent

NEUROLOGICAL injury is a leading cause of permanent 286 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, VOL. 16, NO. 3, JUNE 2008 Optimizing Compliant, Model-Based Robotic Assistance to Promote Neurorehabilitation Eric T. Wolbrecht,

More information

Application of Phased Array Radar Theory to Ultrasonic Linear Array Medical Imaging System

Application of Phased Array Radar Theory to Ultrasonic Linear Array Medical Imaging System Application of Phased Array Radar Theory to Ultrasonic Linear Array Medical Imaging System R. K. Saha, S. Karmakar, S. Saha, M. Roy, S. Sarkar and S.K. Sen Microelectronics Division, Saha Institute of

More information

A Musculoskeletal Model-based Assistance-As-Needed Paradigm for Assistive Robotics

A Musculoskeletal Model-based Assistance-As-Needed Paradigm for Assistive Robotics A Musculoskeletal Model-based Assistance-As-Needed Paradigm for Assistive Robotics Marc G. Carmichael Submitted in fulfillment of the requirement for the degree of Doctor of Philosophy 2013 The Faculty

More information

D. G. Kamper 1,2, W. Z. Rymer 1,2

D. G. Kamper 1,2, W. Z. Rymer 1,2 A BIOMECHANICAL SIMULATION OF THE EFFECT OF THE EXTRINSIC FLEXOR MUSCLES ON FINGER JOINT FLEXION D. G. Kamper 1,2, W. Z. Rymer 1,2 1 Sensory Motor Performance Program, Rehabilitation Institute of Chicago

More information

FEASIBILITY OF EMG-BASED CONTROL OF SHOULDER MUSCLE FNS VIA ARTIFICIAL NEURAL NETWORK

FEASIBILITY OF EMG-BASED CONTROL OF SHOULDER MUSCLE FNS VIA ARTIFICIAL NEURAL NETWORK FEASIBILITY OF EMG-BASED CONTROL OF SHOULDER MUSCLE FNS VIA ARTIFICIAL NEURAL NETWORK R. F. Kirsch 1, P.P. Parikh 1, A.M. Acosta 1, F.C.T. van der Helm 2 1 Department of Biomedical Engineering, Case Western

More information

MODELING OF THE CARDIOVASCULAR SYSTEM AND ITS CONTROL MECHANISMS FOR THE EXERCISE SCENARIO

MODELING OF THE CARDIOVASCULAR SYSTEM AND ITS CONTROL MECHANISMS FOR THE EXERCISE SCENARIO MODELING OF THE CARDIOVASCULAR SYSTEM AND ITS CONTROL MECHANISMS FOR THE EXERCISE SCENARIO PhD Thesis Summary eng. Ana-Maria Dan PhD. Supervisor: Prof. univ. dr. eng. Toma-Leonida Dragomir The objective

More information

4.2 Fatigue at Joint Level versus Fatigue at Muscle Level

4.2 Fatigue at Joint Level versus Fatigue at Muscle Level Chapter 4 Fatigue Model Description 4.1 Introduction The human body is continuously under the influence of external or internal forces. The application of forces may result in beneficial effects (i.e.

More information

Biomechanical Modelling and Animating Human Hand Movements

Biomechanical Modelling and Animating Human Hand Movements Biomechanical Modelling and Animating Human Hand Movements O.A. Kuchar and J.N. Scrimger School of Computer Science, Technical University of Nova Scotia, Halifax, N.S., Canada, B3J 2X4 Abstract A different

More information

Neurobiology guides robotics

Neurobiology guides robotics M. Mahdi Ghazaei Ardakani Neurobiology guides robotics As robotic technology moves toward more anthropomorphic structures with increased complexity, it is reasonable to consider controllers inspired by

More information

A PERCEPTUAL MOTOR CONTROL MODEL BASED ON OUTPUT FEEDBACK ADAPTIVE CONTROL THEORY

A PERCEPTUAL MOTOR CONTROL MODEL BASED ON OUTPUT FEEDBACK ADAPTIVE CONTROL THEORY A PERCEPTUAL MOTOR CONTROL MODEL BASED ON OUTPUT FEEDBACK ADAPTIVE CONTROL THEORY Hirofumi Ohtsuka, Koki Shibasato Department of Electronic Control, Kumamoto National College of Technology 2659-2 Suya,

More information

Active Insulin Infusion Using Fuzzy-Based Closed-loop Control

Active Insulin Infusion Using Fuzzy-Based Closed-loop Control Active Insulin Infusion Using Fuzzy-Based Closed-loop Control Sh. Yasini, M. B. Naghibi-Sistani, A. Karimpour Department of Electrical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran E-mail:

More information

Why Do People Jump the Way They Do?

Why Do People Jump the Way They Do? ARTICLE Why Do People Jump the Way They Do? Maarten F. Bobbert and A. J. Knoek van Soest Institute for Fundamental and Clinical Human Movement Sciences, Vrije Universiteit, Amsterdam, The Netherlands BOBBERT,

More information

ACCURATE NUMERICAL ANALYSIS OF GEAR STRENGTH BASED ON FINITE ELEMENT METHOD

ACCURATE NUMERICAL ANALYSIS OF GEAR STRENGTH BASED ON FINITE ELEMENT METHOD 31 st December 212. Vol. 46 No.2 25-212 JATIT & LLS. All rights reserved. ACCURATE NUMERICAL ANALYSIS OF GEAR STRENGTH BASED ON FINITE ELEMENT METHOD XUEYI LI, CHAOCHAO LI, DAQIAN GENG, SHOUBO JIANG, BINBING

More information

Optimization of Backward Giant Circle Technique on the Asymmetric Bars

Optimization of Backward Giant Circle Technique on the Asymmetric Bars Journal of Applied Biomechanics, 2007, 23, 300-308 2007 Human Kinetics, Inc. Optimization of Backward Giant Circle Technique on the Asymmetric Bars Michael J. Hiley and Maurice R. Yeadon Loughborough University

More information

Design and Implementation study of Remote Home Rehabilitation Training Operating System based on Internet

Design and Implementation study of Remote Home Rehabilitation Training Operating System based on Internet IOP Conference Series: Materials Science and Engineering PAPER OPEN ACCESS Design and Implementation study of Remote Home Rehabilitation Training Operating System based on Internet To cite this article:

More information

What is Kinesiology? Basic Biomechanics. Mechanics

What is Kinesiology? Basic Biomechanics. Mechanics What is Kinesiology? The study of movement, but this definition is too broad Brings together anatomy, physiology, physics, geometry and relates them to human movement Lippert pg 3 Basic Biomechanics the

More information

Nervous System: Spinal Cord and Spinal Nerves (Chapter 13)

Nervous System: Spinal Cord and Spinal Nerves (Chapter 13) Nervous System: Spinal Cord and Spinal Nerves (Chapter 13) Lecture Materials for Amy Warenda Czura, Ph.D. Suffolk County Community College Eastern Campus Primary Sources for figures and content: Marieb,

More information

http://www.diva-portal.org This is the published version of a paper presented at Future Active Safety Technology - Towards zero traffic accidents, FastZero2017, September 18-22, 2017, Nara, Japan. Citation

More information

Muscle Function: Understanding the Unique Characteristics of Muscle. Three types of muscle. Muscle Structure. Cardiac muscle.

Muscle Function: Understanding the Unique Characteristics of Muscle. Three types of muscle. Muscle Structure. Cardiac muscle. : Understanding the Unique Characteristics of Muscle Scott Riewald United States Olympic Committee Three types of muscle Cardiac muscle Involuntary Smooth muscle Involuntary Skeletal muscle Voluntary Involuntary

More information

DESIGN AND IMPLEMENTATION OF A MECHATRONIC SYSTEM FOR LOWER LIMB MEDICAL REHABILITATION

DESIGN AND IMPLEMENTATION OF A MECHATRONIC SYSTEM FOR LOWER LIMB MEDICAL REHABILITATION International Journal of Modern Manufacturing Technologies ISSN 2067 3604, Vol. IV, No. 2 / 2012 17 DESIGN AND IMPLEMENTATION OF A MECHATRONIC SYSTEM FOR LOWER LIMB MEDICAL REHABILITATION Ana-Maria Amancea

More information

Noise Cancellation using Adaptive Filters Algorithms

Noise Cancellation using Adaptive Filters Algorithms Noise Cancellation using Adaptive Filters Algorithms Suman, Poonam Beniwal Department of ECE, OITM, Hisar, bhariasuman13@gmail.com Abstract Active Noise Control (ANC) involves an electro acoustic or electromechanical

More information

Verification of Immediate Effect on the Motor Function of a Plegic Upper Limb after Stroke by using UR-System 2.2

Verification of Immediate Effect on the Motor Function of a Plegic Upper Limb after Stroke by using UR-System 2.2 Verification of Immediate Effect on the Motor Function of a Plegic Upper Limb after Stroke by using UR-System. Hiroki Sugiyama,a, Hitomi Hattori,b, Ryosuke Takeichi,c, Yoshifumi Morita,d, Hirofumi Tanabe,e

More information

Simulation of Tremor on 3-Dimentional Musculoskeletal Model of Wrist Joint and Experimental Verification

Simulation of Tremor on 3-Dimentional Musculoskeletal Model of Wrist Joint and Experimental Verification Simulation of Tremor on 3-Dimentional Musculoskeletal Model of Wrist Joint and Experimental Verification Peng Yao, Dingguo Zhang, Mitsuhiro Hayashibe To cite this version: Peng Yao, Dingguo Zhang, Mitsuhiro

More information

AT89C51 Microcontroller Based Control Model for Hybrid Assistive Limb (Knee )

AT89C51 Microcontroller Based Control Model for Hybrid Assistive Limb (Knee ) From the SelectedWorks of suresh L Winter January 10, 2013 AT89C51 Microcontroller Based Control Model for Hybrid Assistive Limb (Knee ) suresh L Available at: https://works.bepress.com/suresh_l/2/ AT89C51

More information

Skeletal Muscles and Functions

Skeletal Muscles and Functions Skeletal Muscles and Functions Huei-Ming Chai, PT, Ph.D. School of Physical Therapy National Taiwan University Classification of Muscles striated muscles skeletal muscles: voluntary contraction cardiac

More information

Learning Utility for Behavior Acquisition and Intention Inference of Other Agent

Learning Utility for Behavior Acquisition and Intention Inference of Other Agent Learning Utility for Behavior Acquisition and Intention Inference of Other Agent Yasutake Takahashi, Teruyasu Kawamata, and Minoru Asada* Dept. of Adaptive Machine Systems, Graduate School of Engineering,

More information

Neural Facilitation in MRI via Pneumatically Driven, Tele-Operated Systems

Neural Facilitation in MRI via Pneumatically Driven, Tele-Operated Systems Neural Facilitation in MRI via Pneumatically Driven, Tele-Operated Systems Georgia Institute of Technology Milwaukee School of Engineering North Carolina A&T State University Purdue University University

More information

Machines. Design, Modeling and Real-Time Control of Advanced Exercise

Machines. Design, Modeling and Real-Time Control of Advanced Exercise Design, Modeling and Real-Time Control of Advanced Exercise Machines Hanz Richter, PhD, Professor of Mechanical Engineering Control, Robotics and Mechatronics Lab Cleveland State University Background

More information

IMPLEMENTATION AND EVALUATION OF NEUROMUSCULAR CONTROLLERS IN ROBOTIC

IMPLEMENTATION AND EVALUATION OF NEUROMUSCULAR CONTROLLERS IN ROBOTIC ! IMPLEMENTATION AND EVALUATION OF NEUROMUSCULAR CONTROLLERS IN ROBOTIC SYSTEMS PERFORMING COOPERATIVE TASKS WITH HUMANS By DANNY GODBOUT A thesis submitted in partial fulfillment of The requirements for

More information

A hybrid approach for identification of root causes and reliability improvement of a die bonding process a case study

A hybrid approach for identification of root causes and reliability improvement of a die bonding process a case study Reliability Engineering and System Safety 64 (1999) 43 48 A hybrid approach for identification of root causes and reliability improvement of a die bonding process a case study Han-Xiong Li a, *, Ming J.

More information

The Human Machine: Biomechanics in Daily Life.

The Human Machine: Biomechanics in Daily Life. The Human Machine: Biomechanics in Daily Life www.fisiokinesiterapia.biz Biomechanics The study or application of mechanics to biological systems. The study of the forces that act on the body and their

More information

Development of Ultrasound Based Techniques for Measuring Skeletal Muscle Motion

Development of Ultrasound Based Techniques for Measuring Skeletal Muscle Motion Development of Ultrasound Based Techniques for Measuring Skeletal Muscle Motion Jason Silver August 26, 2009 Presentation Outline Introduction Thesis Objectives Mathematical Model and Principles Methods

More information

A Haptic Control Interface for a Motorized Exercise Machine

A Haptic Control Interface for a Motorized Exercise Machine 2008 IEEE International Conference on Robotics and Automation Pasadena, CA, USA, May 19-23, 2008 A Haptic Control Interface for a Motorized Exercise Machine Craig R. Carignan and Jonathan Tang Abstract

More information

Multi-joint limbs permit a flexible response to unpredictable events

Multi-joint limbs permit a flexible response to unpredictable events Exp Brain Res (1997) 117:148±152 Springer-Verlag 1997 RESEARCH NOTE E.M. Robertson R.C. Miall Multi-joint limbs permit a flexible response to unpredictable events Received: 24 March 1997 / Accepted: 7

More information

ECHORD call1 experiment MAAT

ECHORD call1 experiment MAAT ECHORD call1 experiment MAAT Multimodal interfaces to improve therapeutic outcomes in robot-assisted rehabilitation Loredana Zollo, Antonino Salerno, Eugenia Papaleo, Eugenio Guglielmelli (1) Carlos Pérez,

More information

Verification of Passive Power-Assist Device Using Humanoid Robot: Effect on Bending and Twisting Motion

Verification of Passive Power-Assist Device Using Humanoid Robot: Effect on Bending and Twisting Motion 15 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids) November 3-5, 15, Seoul, Korea Verification of Passive Power-Assist Device Using Humanoid Robot: Effect on Bending and Twisting

More information

A Brain Computer Interface System For Auto Piloting Wheelchair

A Brain Computer Interface System For Auto Piloting Wheelchair A Brain Computer Interface System For Auto Piloting Wheelchair Reshmi G, N. Kumaravel & M. Sasikala Centre for Medical Electronics, Dept. of Electronics and Communication Engineering, College of Engineering,

More information

Implementation of Inference Engine in Adaptive Neuro Fuzzy Inference System to Predict and Control the Sugar Level in Diabetic Patient

Implementation of Inference Engine in Adaptive Neuro Fuzzy Inference System to Predict and Control the Sugar Level in Diabetic Patient , ISSN (Print) : 319-8613 Implementation of Inference Engine in Adaptive Neuro Fuzzy Inference System to Predict and Control the Sugar Level in Diabetic Patient M. Mayilvaganan # 1 R. Deepa * # Associate

More information

Electromyographic Correlates of Learning an Internal Model of Reaching Movements

Electromyographic Correlates of Learning an Internal Model of Reaching Movements The Journal of Neuroscience, October 1, 1999, 19(19):8573 8588 Electromyographic Correlates of Learning an Internal Model of Reaching Movements Kurt A. Thoroughman and Reza Shadmehr Department of Biomedical

More information

Cervical reflex Giovanni Ralli. Dipartimento di Organi di Senso, Università di Roma La Sapienza

Cervical reflex Giovanni Ralli. Dipartimento di Organi di Senso, Università di Roma La Sapienza Cervical reflex Giovanni Ralli Dipartimento di Organi di Senso, Università di Roma La Sapienza The development of the neck in vertebrates allows the individual to rotate the head independently of the trunk

More information