Tien-Fu Lu

A three‐DOF compliant micromotion stage with flexure hinges

Micromanipulation has enabled numerous technological breakthroughs in recent years, from advances in biotechnology to microcomponent assembly. Micromotion devices commonly use piezoelectric actuators (PZT) together with compliant mechanisms to provide fine motions with position resolution in the nanometre or even sub‐nanometre range. Many multiple degree of freedom (DOF) micromotion stages have parallel structures due to better stiffness and accuracy than serial structures. This paper presents the development of a three‐DOF compliant micromotion stage with flexure hinges and parallel structure for applications requiring motions in micrometres. The derivation of a simple linear kinematic model of the compliant mechanism is presented and simulation results before and after calibration are compared with results from finite element (FE) modeling and experiments. The position control system, which uses an experimentally determined constant‐Jacobian, and its performance are also presented and discussed.

Comparing Insect-Inspired Chemical Plume Tracking Algorithms Using a Mobile Robot

Four chemical plume-tracking algorithms have been compared using a mobile robot. These algorithms are based upon hypotheses proposed to explain the plume-tracking behavior of flying insects. They all use information from a wind sensor and a single chemical sensor to determine how the agent should move to locate the source of the chemical plume. The performance of the robot using each of the algorithms was tested in a wind tunnel under a range of wind speeds (0.55, 0.95, and 1.4 m/s) using a model chemical (ionized air). The robot was capable of tracking the ion plume to its source effectively with each algorithm, having an overall success rate of over 85%. The simplest implemented algorithm, surge anemotaxis, was found to be the fastest. However, the shape of the tracking paths observed indicated that this simple algorithm may not explain the plume-tracking behavior of certain insects as well as the other algorithms tested. Further tests are required to see if the surge anemotaxis algorithm remains the most efficient under more realistic wind conditions.

An on-line relative position and orientation error calibration methodology for workcell robot operations

Abstract Relative position and orientation inaccuracy always exists between a robot and the equipment with which it operates, especially in batch-type production cells that are subjected to dynamic changes. This inaccuracy causes robot relative positioning errors, and may even result in operation failure if the off-line programmed moving path is implemented without adjustment. To make use of off-line programming and simulation tools, an on-line calibration methodology for robot relative positioning inaccuracy was developed in this study. This methodology eliminates the need for time-consuming off-line calibrations relying on accurate models and expensive devices. An industrial robot system was enabled to detect and compensate automatically for relative positioning errors by incorporating a vision system, a 3-D force/torque sensor, and control strategies involving neural networks. The experimental results showed that this methodology is valid and robust in calibrating the relative position and orientation errors automatically without the need for mathematical models and complex off-line calibration procedures for model parameters. Consequently, batch-type production cells would be more flexible, adaptable and intelligent in accommodating dynamic workcell changes with less human effort.

Loop closure theory in deriving linear and simple kinematic model for a 3 DOF parallel micromanipulator

Various types of micro-motion devices have been developed in the past decade for applications including the manipulation of cells in micro-surgery and the assembly of micro-chips in micro-assembly industries. Most of the micro-motion devices are designed using the compliant mechanism concept, where the devices gain their motions through deflections. In addition, closed-loop parallel structures are normally adopted due to better stiffness and accuracy compared to the serial structures. However, the forward kinematics of parallel structures are complex and non-linear; to solve these equations, a numerical iteration technique has to be employed. This iteration process will increase computational time, which is highly undesirable. This paper presents a method of deriving a simple, linear and yet effective kinematic model based on the loop closure theory and the concept of the pseudo-rigid-body model. This method is illustrated with a 3 DOF (degree-of-freedom) micro-motion device. The results of this linear method are compared with a full kinematic model for the same micro-motion system. It is proved that the derived kinematic model in this paper is accurate and the methodology proposed is effective. The static model of the micro-motion device will also be presented. The uncoupling property of the micro-motion systems, based on the static model, will be briefly discussed.

The advancement of an autonomous underwater vehicle (AUV) technology

An autonomous underwater vehicle (AUV) test-bed was jointly upgraded by DSTO and the University of Adelaide as an introduction of advanced technologies into its system. This AUV will serve as a platform to support further research dealing with underwater robotic operations. The group working on this project incorporated off-the-shelf hardware together with well-developed control algorithms to build an efficient AUV embedded system. The presented system consists of five main modules including command, navigation, communication, and instrumentation (internal and external sensors) and machine-vision. This paper presents an overview of the AUV development process which covers project stages, system development, system operation as well as test results from experimental trials which showcased the capability and functionality of the AUV.

Preparation for turning a Bucket Wheel Reclaimer into a robotic arm

This paper presents the ground work prepared for turning a Bucket Wheel Reclaimer (BWR) into a robotic arm for operation automation. BWR has been widely used in dealing with bulk materials by removing them from the conveyor and stack them as well as to load stacked materials onto the in-feed conveyor, such as in mining industry. Generally speaking, current BWRs are manually operated, remotely operated, or automated to simply follow predefined trajectory patterns. In order to automatically generate trajectory and manoeuvre BWR to specific stockpiles and locations to reclaim materials for the combined product grade required by customers, knowledge of the product quality distribution among stockpiles and within a stockpile is essential. In addition, a converted BWR system like a robotic arm, which can be automated to manoeuvre according to the needs, is required. This paper focuses on the preparation for converting BWR only and presents the derivation of its forward/inverse kinematics, Jacobian and dynamics. A model based controller is developed as well to move around the converted robotic arm. Simulations are conducted using Matlab/Simulink. The results demonstrate satisfactory performance and prove the initial stage of preparation is successfully completed.

Effectiveness of Insect-Inspired Chemical Plume-Tracking Algorithms in a Shifting Wind Field

Tracking a plume of chemical back to its source is made difficult due to the complexity of plume structure caused by turbulence and shifts in the prevailing wind direction. Insects overcome this problem using forms of anemotaxis, which involve traveling upwind when an attractive chemical is perceived. In this study, two series of insect-inspired plume-tracking algorithms were implemented on a mobile robot and their performance compared under changing wind conditions in a wind tunnel. The robot was capable of sensing wind velocity and the level of a plume of ions. Ion sensors respond and recover far more rapidly than do conventional chemical sensors, so the substitution of an ion plume for a chemical plume allowed the algorithms to be implemented with rapid responses to changing plume concentration. The addition of a specific behavioral response to a wind shift decreased the time taken for the robot to find the plume source in the event of a wind shift. Increased speed came with only a minor drop in the success rate of the searching. Anemotaxis is an effective approach to chemical plume tracking with robots. The performance of these simple algorithms can be improved by modest increases in the complexity of the algorithms.

A simple and efficient dynamic modeling method for compliant micropositioning mechanisms using flexure hinges

In this paper we consider the dynamic modelling of compliant micropositioning mechanisms using flexure hinges. A simple modelling method is presented that is particularly useful for modelling parallel micropositioning mechanisms. This method is based upon linearisation of the geometric constraint equations of the compliant mechanism. This results in a linear kinematic model, a constant Jacobian and linear dynamic model. To demonstrate the computational simplicity of this methodology it is applied to a four-bar linkage using flexure hinges. Comparisons are made between the simple dynamic model and a complete non-linear model derived using the Lagrangian method. The investigation reveals that this new model is accurate yet computationally efficient and simple to use. The method is then further applied to a parallel 3-degree of freedom (dof) mechanism. It is shown that the method can be simply applied to this more complex parallel mechanism. A dynamic model of this mechanism is desired for use in optimal design and for controller design.

Neural-network-based 3D force/torque sensor calibration for robot applications

Abstract The integration of high-performance sensor systems with a robot is essential to enhance the “intelligent” capability of the robot, while force/torque sensors are especially important sources of feedback in robot applications, for proper monitoring, analysis and force/motion control. Being integrated with robots, force/torque sensors may suffer from non-linearity and various forms of uncertainty, resulting not only from the sensors themselves, but also from the workcells they integrate; therefore the conventional least-squares method for force/torque sensor calibration is unable to interpret the relationships between sensor readings and their represented outputs accurately enough for some applications. This paper therefore presents the development of a new approach, using neural networks for the efficient and accurate calibration of the F/T sensors integrated with robots. Apart from the neural-network-based force/torque sensor-calibration methodology, this paper also presents the calibration implementation using both the least-square method and the proposed method, and a comparison and discussion of the calibration results. These results show that the proposed neural-network-based calibration method is more efficient and accurate, which verifies the adaptability and applicability of this method to nonlinear and dynamic robot systems.

Optimization of reclaiming voxels for quality grade target with reclaimer minimum movement

This paper presents the first part of study on the proposed automatic optimal reclaiming system. The proposed system focuses on how the reclaimer reclaims iron ore from stockpiles to fulfill quality grade target aiming for minimum reclaimer overall movement. For example, iron ore can be excavated from open-cut pits, stacked onto stockpiles and reclaimed later using bucket wheel reclaimer (BWR) for export to customers. Stockpiles here are represented by combinations of voxels each possessing certain quality distribution. The supply of iron ore closest to the target grade with lower operation cost is the main concern for the ore producer. Currently, BWRs are operated manually to meet quality grade demand involving the checking of quality periodically during reclaiming operation. In result, there is excessive movement of the large machine leading to higher energy consumption. In this paper, the combination of voxels is identified from available stockpile voxels considering BWR minimum movement to fulfill quality and quantity demand of iron ore. Achieving the desired grade with maximum efficiency will result in increasing productive rates and less impact on global warming by significantly reducing energy consumption.