ADVANCES IN BlOROBOTIC VlSlON

The Australian National University Biorobotic Vision Group has described its innovative researches and developments in the field of robotic vision. The group has been researching the principles by which insects see, control flights and navigate, in order to transfer their findings to their work in robotic vision. They believe this work will help them in their aim to devise algorithms for machine vision and autonomous, visually guided robots. In particular, this work is in reference to using optic flow to measure image motion.

The researchers have already successfully developed a hover controller that can hover over an arbitary set of landmarks with no manual intervention.They have also designed an autonomous helicopter that will hover and fly without using GPS.

This paper addresses the problem of global output feedback, link position tracking control of robot manipulators. Specifically, a robust, Lyapunov-based controller is designed to ensure that the link position tracking error is globally uniformly ultimately bounded despite the fact that only link position measurements are available in the presence of incomplete model information (i.e., parametric uncertainty and additive bounded disturbances).

The purpose of this paper is in duplicate to present computer simulation results of concurrent grasp and object manipulation by a pair of three degrees of freedom (3-dof) robot fingers with rigid hemispherical finger-ends that induce rolling contacts with an object and propose a guidance of gain tuning. Although the existence of a class of sensory feedback signals that realize stable grasp and orientation control of the object concurrently has been shown theoretically, the problem of tuning of their feedback gains has not yet been solved. This paper proposes a guideline for tuning sensory feedback gains by deriving a relationship between the object mass and damping coefficients of finger motions through analyzing the overall fingers-object dynamics and taking into account the well-known force/velocity characteristics of human muscle in muscle physiology.

It is reported by the British Aerospace (BAE Systems) that their researchers have developed a revolutionary airborne computer system that, they claim, can land aircraft safely without the need of human air traffic controllers. The new system will allow a pilot to determine an aircraft's landing by pressing a button in the cockpit to initiate the computer system.

This paper analyses the concept of velocity isotropy for Parallel Mechanisms with Actuation Redundancy (PMAR). The limits of classical indices based on the Jacobean matrix condition number are shown. It is proposed to use either the largest ellipsoid included in the operational polytop, or the operational polytop itself, as better representations of a PMAR capabilities. The polytop is studied because it is actually the accurate representation of a machine capability, while the largest ellipsoid remains similar to the classical tool roboticists are dealing with for decades. Velocity performance indices are proposed, and the ways to compute them efficiently are given.

The statistics provided by the United Nations Economic Commission (ECE) and the International Federation of Robotics (IFR) supply us with an up-to-date insight into the world robot market. This joint report tells us a great deal about the technological evolution of industrial robots since their introduction at the end of the 1960s. We know that real prices have declined, and the performance of robots is continuously improving. Robots are now in use in a wide range of applications in countries worldwide.

This paper deals with the controllability of biologically inspired snake-like mechanisms. Motion equations of the generic snake-like mechanism and investigations on its controllability lead to extremely complicated mathematical representations. The authors establish a physical analogy between the undulatory snake locomotion and object grasping. Based on this analogy, they propose a controllability criterion concerning the motion of the mobile mechanism. Results of dynamic simulations of the controlled system systematically point to the validity of the analogy-based controllability criterion. The purpose of this work is to achieve an understanding of the undulatory motion which constitutes the basis of a biologically motivated experimental setup.

A two-straight-legged walking mechanism with flat feet is designed and built to study the passive dynamic gait. It is shown that the mechanism having flat feet can exhibit passive dynamic walking as those with curved feet, but the walking efficiency is significantly lower. It is also shown that the balancing mass and its orientation are effective for controlling side-to-side rocking and yaw, which have significant effects on steady walking. The effects of various parameters on the gait patterns are also studied. lt is shown that changes in the ramp angle have the most dominant effect on the gait pattern as compared with the changes in the hip mass, ramp surface friction and size of the flat feet. More specifically, as the ramp angle increases, the step length increases while the range of the side-to side rocking angle decreases and the step length dictates the walking speed and the gravitational power. Another finding, is that adding a hip mass improves the walking efficiency by allowing the mechanism to walk on a flatter ramp. This research enables us to gain a better understanding of the mechanics of walking. Such an understanding will have a direct impact on better design of prostheses and on the active control aspects of bipedal robots.

We describe the design and implementation of RSTATION, an object-oriented, modular robot simulator with hierarchical analysis capabilities. Modularity is achieved via the features of design encapsulation and enables grouping a set of interconnected components into a single component and dividing the robot system into several sets of subordinate modules recursively. By careful construction of the data types and classes, RSTATION allows for hierarchical simulation of the kinematics, and the dynamics at three levels: considering only main links (high-level), using simplified models including dynamic properties of transmission elements (intermediate level), and taking into account the detailed kinematics and dynamics of transmission elements (low-level). Submodules can be set to different resolution during a single simulation. The data types and classes also exploit a recent set of coordinate invariant robot analysis algorithms based on modern screw theory. Central to the low-level dynamic analysis capability is an algorithm for systematically extracting the constraint equations for general gearing systems. The various features of RSTATION are illustrated with a detailed case study of a commercial industrial robot.

This paper proposes a novel path planning algorithm of 3-D articulated robots with moving bases based on a generalized potential field model. The approach computes, similar to that done in electrostatics, repulsive forces and torques between charged objects. A collision-free path can be obtained by locally adjusting the robot configuration to search for minimum potential configurations using these forces and torques. The proposed approach is efficient since these potential gradients are analytically tractable. In order to speedup the computation, a sequential planning strategy is adopted. Simulation results show that the proposed algorithm works well, in terms of collision avoidance and computation efficiency.

The purpose of the present study is to assess the effect of an auxiliary lateral image and display devices on manipulation performance in a virtual reality-based hand rehabilitation system. The system consists of a personal computer, a tracker, a data glove, and a display device. For this study, a projector, a monitor, and a head-mounted display were respectively used as the display devices to present three-dimensional virtual environments. Twelve volunteers were recruited to take a pick-and-place procedure at different levels of difficulty. Task time and collision frequency were the parameters used to evaluate the manipulation performance. It was found that the presence of an auxiliary lateral image was a significant factor only for the performance of the projector group and the monitor group. In addition, no statistically significant difference was found in the comparison between the projector group and the monitor group.

Using a coarse-and-fine actuator combination (dual stage system), a new design of the three degree-of-freedom (DOF) micro parallel positioning platform with high mobility, high accuracy, and a large working space is proposed. To achieve these three DOFs and implement the dual stage system, there are six possible architectures for the coarse and fine actuators, respectively. This paper is organized in two parts. Part I treats the kinematic analysis of each architecture and the problem of selecting the correct coarse actuator architecture. Inverse kinematics and Jacobian matrices for six types of coarse actuator architectures are derived and one proper coarse actuator architecture is selected based on the mobility (rotational capability) analysis, condition number evaluation of the Jacobian matrix, and manufacturability consideration. Part II on real machine design will follow in the next issue of Robotica.

A method is proposed to solve the inverse kinematics and control problems of robot control systems using a cerebellar model articulation controller neural network combined with a genetic algorithm. Computer simulations and experiments with a 7-DOF redundant modular manipulator have demonstrated the effectiveness of the proposed method.

Although numerous sophisticated nonlinear control algorithms exist in literature, it is still state of the art to use simple linear joint controllers in industrial robotic systems. Most nonlinear concepts are based on a more or less accurate inverse model of the robot. In this paper a forward-model-based control system, the so-called Model Following Control (MFC), for robot manipulators is presented. Its theoretical basics and its concept are explained. The quality and the applicability of the MFC control concept has been analyzed in many experiments. The MFC system is compared with classical linear controllers and nonlinear feedforward controllers with respect to robustness. Qualitative as well as quantitative results are presented and discussed.

This paper deals with the design and development of a new perception system for controlling a set of ultrasonic transducers working at two different frequencies generated by a digital signal processor (DSP). Data gathering speed is increased by the simultaneous firing of sonars without crosstalk problems. The software and hardware implementation are presented as well as the results from the calibration process and the integration on board the mobile robot RAM-2. Finally, a demonstration of the use of the proposed equipment to wall-following via a fuzzy controller is described.

Autonomous navigation of an indoor mobile robot, using the global ultrasonic system, is presented in this paper. Since the trajectory error of the dead-reckoning navigation increases significantly with time and distance, the autonomous navigation system of a mobile robot requires self-localization capa-bility in order to compensate for trajectory error. The global ultrasonic system, consisting of four ultrasonic generators fixed at a priori known positions in the work space and two receivers mounted on the mobile robot, has a similar structure to the well-known satellite GPS(Global Positioning System), which is used for the localization of ground vehicles. The EKF (Extended Kalman Filter) algorithm is utilized for self-localization and autonomous navigation, based on the self-localization algorithm is verified by experiments performed in this study. Since the self-localization algorithm is efficient and fast, it is appropriate for an embedded controller of a mobile robot.

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

This paper presents a control law for the tracking of a cyclic reference path by an under-actuated biped robot. The robot studied is a five-link planar biped. The degree of under-actuation is one during the single support phase. The control law is defined in such a way that only the geometric evolution of the biped configuration is controlled, but not the temporal evolution. To achieve this objective, we consider a parametrized control. When a joint path is given, a five degree of freedom biped in single support becomes similar to a one degree of freedom inverted pendulum. The temporal evolution during the geometric tracking is completely defined and can be analyzed through the study of a model with one degree of freedom. Simple analytical conditions, which guarantee the existence of a cyclic motion and the convergence towards this motion, are deduced. These conditions are defined on the reference trajectory path. The analytical considerations are illustrated with some simulation results.

This paper deseribes the control of multi-arm co-operating manipulator systems handling a common object. Inverse dynamics controllers with motive force compensation are developed for the co-operating fixed-base, free-floating, and free-flying space manipulator systems. Further, the relative performances of all the three configurations are compared on two tracking problems.

This paper presents two frontal plane algorithms for 3D dynamic bipedal walking. One of which is based on the notion of symmetry and the other uses reinforcement learning algorithm to learn the lateral foot placement. The algorithms are combined with a sagittal plane algorithm and successfully applied to a simulated 3D bipedal robot to achieve level ground walking. The simulation results showed that the choice of the local control law for the stance-ankle roll joint could significantly affect the performance of the frontal plane algorithms.