We propose a general and practical planning framework for generating 3-D collision-free motions that take complex robot dynamics into account. The framework consists of two stages that are applied iteratively. In the first stage, a collision-free path is obtained through efficient geometric and kinematic sampling-based motion planning. In the second stage, the path is transformed into dynamically executable robot trajectories by dedicated dynamic motion generators. In the proposed iterative method, those dynamic trajectories are sent back again to the first stage to check for collisions. Depending on the application, temporal or spatial reshaping methods are used to treat detected collisions. Temporal reshaping adjusts the velocity, whereas spatial reshaping deforms the path itself. We demonstrate the effectiveness of the proposed method through examples of a space manipulator with highly nonlinear dynamics and a humanoid robot executing dynamic manipulation and locomotion at the same time.

Catching the moving object by a robot manipulator when controlling the robot via the Internet is considered at this paper. The object was the rod on bifilar suspension, performing complicated oscillations. Task statement and capture algorithms are presented. Shared autonomy control was used to implement the capture. Operator forms initial motion of the rod, chooses desired point for capture on the rod and capture instant. Final operation is performed automatically. Operator plans capture strategy using 3D on-line virtual model of the robot and object in its current positions. Capture operation is performed using motion prediction algorithm, based on the object’s motion model. Experimental test-bed, remote control interface are presented. The particulars of the system, the experiments undertaken, current issues, and directions of the future work are discussed. Video film will demonstrate results of the work.

This paper deals with motion planning algorithms for the large space robot manipulators with complicated dynamic behavior. We propose two "two-stage" iterative algorithms, which provide collision-free robot motion taking into account robot's dynamics. The approach is based on new efficient methods for robot manipulator dynamics simulation and probabilistic methods for motion planning in highly cluttered environments. The algorithms are applicable for the robot manipulators of general class with arbitrary kinematics and dynamics parameters. We have demonstrated the approach for a particular task of servicing the satellite by a large space manipulator. This task is one of the most challenging since large space manipulators have extremely complicated dynamic behavior caused by elasticity of their structure, huge payloads they work with and zero-gravity conditions. Experiments involving a 15.5 meters long manipulator carrying a satellite inside a space shuttle with clearance less than 3 cm are presented. Several movies demonstrate the results.

A virtual control environment for robot teleoperation via the Internet is presented. It comprises a Java3D-based real-time virtual representation of the robot and worksite, and uses a graphic panel, an environment for remote robot programming, and a dataglove with a 6D position tracker as the control interfaces. The use of virtual reality (VR) techniques for Internet teleoperation allow: (1) time delays inherent in IP networks to be suppressed and (2) the operator's work to be simplified and accelerated, compared to methods that use delayed TV images. The system realisation, with its use of open technologies Java, Java3D and 3-tier client/server architecture, provides portability among different computer platforms and types of robots. The efficiency of the VR-based methods developed has been verified for slow communication rates (0.1-0.5 KB/sec), where TV-based control methods are inapplicable. VR systems have been developed for the WWW-based control of the PUMA and CRS industrial robot manipulators. The particulars of these systems, the experiments undertaken, current issues, and directions of future work are presented.

A system for WWW-based remote robot tele-operation is presented. It addresses important problems of Internet telerobotics such as "how do we make the system reliable?", "how do we make the system fast?", and finally, "how do we make the system open and portable?". The recent open technologies Java and Java3D are used for developing the system. Key features of the system are: (1) the Java3D based "live" virtual representation of the real robot and its environment, and (2) the use of a powerful tool for the remote programming of the robot. Possible applications of the teleoperation tools developed and methods within VR are also briefly discussed.

The paper presents the results in controlling the robot manipulator handling the moving object of various types. Two series of experiments were carried out using the rod swinging on the bifilar suspension, and the tennis balls swinging as spherical pendulums. The main part of the experiments with the rod was devoted to finding efficient ways for its fast and reliable capturing. The experiments with the ball were focused on modifying the motion of the ball in such a way for its orbit has the desirable parameters. In particular, the ball was directed to reach the motionless target placed inside the working space, or perform a rendezvous with another ball having an independent free motion. The TV camera supplied the computer with the visual data of the scene. The dynamical models of the motion of the objects were used for statistical processing the visual data and determining the parameters of the motion. These parameters were used for controlling the robot.

The paper contents description of the experiments on capturing the moving object by means of a robot manipulator RM-01 (like PUMA 560) with vision system. The object to be grasped is the rod on a bifilar suspension. The upper ends of the fibers are attached to the stable beam which holds a TV camera connected to a computer. Data processing results in the coordinates of the end points of the rod. The algorithm developed gives the possibility to obtain up to 25 pairs of end points per second. The data are subjected to statistical analysis taking into account the dynamic model of the rod motion. The results obtained for the 0.5 s interval allow us to calculate the initial conditions of the rod motion and to predict the motion to several seconds ahead. Analysis of the predicted motion allows us to determine the proper moment of capture, and the position of the manipulator grip at that moment. The capture is implemented in 2-5 s after the start of the process, within 1-2 periods of longitudinal and lateral modes of the rod oscillation.