Ming-Yen Li

Bio: Ming-Yen Li is an academic researcher from National Taipei University of Technology. The author has contributed to research in topics: Obstacle & Factory (object-oriented programming). The author has an hindex of 1, co-authored 1 publications receiving 7 citations.

Motion planning with obstacle avoidance of an UR3 robot using charge system search

Abstract: For a cyber-physical system (CPS) of a future intelligent factory, a robotic manipulator is requested to co-work with human efficiently and safely in an environment with flexible arrangements. Therefore, an autonomous path planning of robotic manipulator is the most necessary issue to be resolved for the factory automation. For the robotic manipulator, optimizations and artificial intelligence (AI) methods are widely used to investigate the autonomous dynamic path-planning tasks with obstacle avoidance. Among these methods, the Rapidly Exploring Random Tree (RRT) algorithm has been widely used in path planning for a complex environment, because the RRT algorithm has the advantages of perfect expansion, probability completeness, and fast exploring speed. However, for some practical cases, the existing RRT algorithm may obtain a discontinuous solution of the angular trajectory. To solve the above problem, we studied a particle swarm optimization with the charge search system (CSS) to find the optimal path planning with obstacle avoidance. The steps of the proposed method are mentioned as follows: (1) establish the configuration space with the obstacle regions, (2) formulate the motion planning with obstacle using the CSS method and (3) use the PSO method to solve the path planning problem. Finally, the simulation of the path-planning task with obstacle avoidance is visually illustrated using the software RoboDK and the proposed method is implemented by the real-time experiments of the UR3 robot.

Human machine interface based on virtual reality for programming industrial robots

Abstract: In the paper the new approach to the automated generation of the industrial robot tool trajectory was developed. This approach is based on the use of the operator interface implemented in the virtual reality environment. This interface displays a three-dimensional model of the industrial robot working area, obtained by means of the computer vision system. This model is used to specify the tool trajectory. The proposed interface allows to simplify and accelerate the industrial robots programming for perform technological operations in a working environment.

Forward and Inverse Kinematics Demonstration using RoboDK and C

Abstract: Purpose: Robot researchers need a simulator to understand better the algorithm on path planning, arm movement, and many more. They need a good simulator. RoboDK is an excellent simulator to fulfill the research work. It has calibration facilities, so it is industrial-grade software. Its forward and inverse kinematics accuracy is better than any competing software. The main advantage is all robots under one IDE. When we use an industrial robot, and we must use their software environment to operate the robot. But the RoboDK covers most of the robots and runs under one roof. And we need to learn only one IDE. The RoboDK online library is full of the standard robot. And all robot’s operation procedure is the same. So, the learning curve of new robots is easy. It is easy to simulate, and it can connect with a practical robot to execute the task. Using this software, we can quickly create digital twins for the industry. Now we think about control the robot from our application. When we use to control the robot from an external environment or remote software, we need the use the API to control the robot. Here we will see how easily we can operate the robot from our custom application. We adopted RoboDK C# API and integrated it into Visual studio using a User interface to control the robot movement. Keeping this research as a reference, the robotic arm researcher can add value to their research. Our primary purpose is to shorten the learning curve to integrate the RoboDK with their custom application. Design/Methodology/Approach: Taking the RoboDK C# API they provided, we customized it according to our purpose with minimal components. After developing a graphical user interface, we interact through API. Then, opening both RoboDK IDE and C# application, we can send the End effector position using the sliding movement. Findings/Result: After our research, we found that RoboDK is a good IDE for our research on the robotics arm. We can easily integrate the C# API they provided with our custom application for research purposes. Originality/Value: If we want to test robotic arm movement in the simulator, we need an excellent simulator like RoboDK. Integrating the RoboDK C# API is a little bit time-consuming. Using our approach, the researcher can continue their research in a minimal period. And find adequate information here to integrate easily into their project. Paper Type: Simulation-based Research.

Comparison of Multiple Reinforcement Learning and Deep Reinforcement Learning Methods for the Task Aimed at Achieving the Goal

Abstract: Reinforcement Learning (RL) and Deep Reinforcement Learning (DRL) methods are a promising approach to solving complex tasks in the real world with physical robots. In this paper, we compare several reinforcement learning (Q-Learning, SARSA) and deep reinforcement learning (Deep Q-Network, Deep Sarsa) methods for a task aimed at achieving a specific goal using robotics arm UR3. The main optimization problem of this experiment is to find the best solution for each RL/DRL scenario and minimize the Euclidean distance accuracy error and smooth the resulting path by the Bezier spline method. The simulation and real word applications are controlled by the Robot Operating System (ROS). The learning environment is implemented using the OpenAI Gym library which uses the RVIZ simulation tool and the Gazebo 3D modeling tool for dynamics and kinematics.

An analysis of automation scenarios in an aerospace environmental testing facility

Abstract: This paper presents a real-life example of using industrial robots to automate the testing procedures of aerospace products. Three different scenarios are proposed as potential solutions to allow a Mobile Manipulator to help automate the environmental testing process. An overview of different simulation and Off Line Programming (OLP) software packages are presented and discussed as methods to test and validate the scenarios. A framework is presented consisting of a product data model to be used for the identification of the different products and a unified modelling system is introduced to allow information exchange at the operational level. The focus of this paper and future work is towards integrating all the elements of the proposed framework into a human machine interface to create a collaborative process involving human, robots, machines and products.

A dual-arm manipulation strategy using position/force errors and Kalman filter:

Abstract: This paper presents a new coordination manipulation strategy for a custom-made dual-arm robot. With master and slave coordination infrastructure, both spatial relation and sense of touch are consid...