Jay Krishnasamy

Neural Netowrk Based Fault Diagnostics of Industrial Robots using Wavelt Multi-Resolution Analysis

A multi-resolution wavelet analysis coupled with a neural network based approach is applied in the problem of fault diagnostics of industrial robots. The multi-resolution analysis implements discrete wavelet transforms with filters and decomposes the signal in various levels. The approximate and detailed coefficients of the decomposed signals are then used for training a feedforward neural network whose output determines the state (faulty or normal) of the robot. The neural network classifier was then implemented and monitored in a Matlab-Simulink environment using a state-flow model. Validation of the method was performed offline using experimental data obtained from an industrial robot manipulator used in the semi-conductor industry.

Fault Diagnostics of Industrial Robots Using Support Vector Machines and Discrete Wavelet Transforms

In this paper we address the problem of fault diagnostics in industrial robots. The goal was to develop a method that automatically, accurately and in a generic way could identify and classify faults once they occur for any type of industrial robot used. Although a large number of diagnosis methods and relevant applications for industrial equipment already exist, the current research in the area of fault diagnosis of industrial robotic manipulators is rather poor, due to the large variability of faults, the unsteady and non-uniform operating conditions, the small amount of sensors used in industrial manipulators and the rather limited time records of the equipment. These restrictions present key challenges of the current research to be undertaken. In this paper we present a novel approach to perform fault diagnostics of industrial robotic systems using Support Vector Machines (SVM) and Discrete Wavelet Transform based feature extraction. Experimental results are obtained from an industrial manipulator used in the semi-conductor industry.Copyright

Positioning Repeatability of Robotic Systems With Synchronous Belt Drives

Substrate-handling robots for pick-place operations in semiconductor manufacturing applications are subject to strict substrate placement repeatability specifications. It has been observed that the placement locations at a given workstation tend to exhibit distinct clusters, each of which can be associated with another workstation accessed by the robot in the past, resulting in an undesirable increase of the overall placement repeatability range. In the present paper, this memory-like repeatability phenomenon is studied, and attributed to multistage synchronous belt drives, which are utilized to transmit motion from centralized motors to individual links and end-effectors of the robot arms. The phenomenon is investigated experimentally, and simulated using a simplified lump-parameter model. The effects of selected belt drive design parameters are examined, and the results are utilized to improve the positioning repeatability performance of a typical substrate-handling robot.Copyright

Reduced-Complexity Dual-Arm Robotic Manipulator for Compact Substrate-Handling Platforms

The design of a reduced-complexity dual-arm robotic manipulator for compact substrate handling platforms is presented. The manipulator utilizes a common component, which forms the left upper-arm link and right upper-arm link of the robot arm, and two forearm links, each of which carries an end-effector and can move substrates in the radial as well as circumferential directions. A unique feature of the design is that a novel transmission mechanism is used to couple a single drive motor to the two forearm links. The advantage of using such a transmission mechanism is that one less motor is required to achieve the desired motion. This article outlines the concept of the robotic manipulator and the transmission mechanism, presents a kinematic and dynamic model of the combined system, and discusses a design methodology so as to satisfy the motion requirements in a compact substrate-handling platform. The feasibility of the proposed concept is demonstrated on a fully functional prototype.Copyright

Modeling of the Axial and Tilt Stiffness Between a Magnet Track and a Ferromagnetic Backing

This paper consists of modeling and analysis of the axial and radial forces as well as axial and tilt stiffness between a magnetic track and a ferromagnetic backing. The proposed model utilizes the concept of Maxwell stress tensor and simplified reluctance models in order to predict the magnetic interactions between the rotor and the stator. In other to verify the model developed, finite element models and experiments are performed for the case of a linear magnetic track. The results can be applied to the case of rotary tracks without loss of generality. Applications of this work can be used in the design of magnetic bearings or servo motors with ferromagnetic stators.Copyright

Design Optimization of an Actuation Mechanism for a Dual-Arm Robotic Manipulator

The design optimization of an actuation mechanism for a dual-arm scara-type robotic manipulator is presented. The manipulator is to be used in a vacuum environment for wafer handling applications in the semi-conductor manufacturing industry. The actuation mechanism consists of a pivoting platform that serves as the common driver link for two four-bar mechanisms each of which drives a scara-type arm. Each of the scara-type arms has a substrate carrying end-effector to pick from or place substrates on a process module. When the pivoting platform is in its neutral position, both of the arms are retracted. When the platform swings to one side, the arm on that side extends while the arm on the other side remains close to retracted position. The actuation mechanism is unique in that it uses just one motor to control the extension of both arms in contrast to the conventional design where each arm requires a motor of its own. However, the design needs to be optimized in order to minimize the effects of kinematic coupling between the two arms and, at the same time, keep the motor torque requirements and encoder resolution requirements to within practical limits. In this article, the relationships between the link lengths of the actuation mechanism, the kinematic coupling between the two arms, maximum encoder resolution requirements and motor torque requirements are described. These relationships are presented in the context of motion profiles characterized by limits on substrate acceleration — a typical requirement in wafer handling applications in vacuum. A methodology for determining link lengths to minimize motor torques and kinematic coupling between arms is presented.Copyright