No abstract

Sheet-piling processes pose the problem of guiding the pile to follow a fixed trajectory in order to drive it correctly into the ground. Since the complete system consists of the supporting excavator boom with hydraulic actuators, the gripper, and the vibratory unit, the large number of system variables implies the use of servo control to allow a human operator to handle the operation. An exact inverse kinematic transformation from Cartesian workspace to boom joint space variables allows the pile penetration to be guided, keeping the valve input of the main boom actuator as the free parameter, while the remaining actuators are governed automatically by the control computer. The penetration record of the pile differs based on whether a constant valve input is used in semiautomatic steering mode or, alternatively, the desired penetration speed is governing the valve input in a feedback loop in full-automatic steering mode. A numerical model representing an actual industrial system has been developed. Variable results from computer simulations have led the manufacturer to realize the steering system. It has been verified in real piling tests that the steering system may be tuned to be stable and fast enough for practical working conditions. In particular, the performance of trajectory control in terms of tracking error is much better than in the conventional manual steering method, as expected.

No abstract

This paper presents a generic method for generating joint trajectories for robotic manipulators with collision avoidance capability. The coordinate motion control system of the heavy-duty hydraulic manipulator resolves joint references so that a goal pose can be reached in real-time without any collisions. The control system checks whether any part of the manipulator is at risk of colliding with itself, with other manipulators, or with environmental obstacles. If there is a risk of collision, then the collision server searches the points where the collision is about to occur and calculates the shortest distance between the colliding objects. The collision server retains static and dynamic point clouds, and it uses point cloud data to calculate the shortest distance between the colliding objects. The point clouds on the server are kept up to date with the manipulators’ joint sensors and an external surveillance system. During coordinated motion control, the joint trajectories of the hydraulic manipulator are modified so that collisions can be avoided, while at the same time, the trajectory of the end-effector maintains its initial trajectory if possible. Results are given for a seven degrees of freedom redundant hydraulic manipulator to demonstrate the capability of this collision avoidance control system.

Hydraulic elevating platforms are commonly used machinery in assembling outside covers to buildings, washing windows etc. Evacuation of people from high places as well as fire-fighting are also well known service areas of elevating platforms. A team of researchers, designers, and end-users has introduced a concept of a man-in-the-loop simulator to be used in operator training for time-critical and accurate boom maneuvers. The hardware consists of a boom platform mounted on a 3d Stewart platform. Virtual engineering software is used to visualize the working environment on wall screens while a real-time simulation model transforms large boom movements to produce restricted motion in the Stewart mechanism.