About Instruments Today No. 231


Adhere to the Principles of “Start Small and Think Big” Wherever You Go, and Will Lead You to Obtain Fruitful Achievements

Claire Lin

Special Issue: Intelligent Robot and Automation

Topic Introduction of “Intelligent Robot and Automation”

Liang-Chia Chen

Utilizing Visual Deep Learning for Real-time Object Detections of Humanoid Robot Soccer Games

Ken-Fun Lin, Chung-Hsien Kuo

RoboCup is one of the most important humanoid robotic competitions in the world. The development of humanoid soccer robot considers not only the bipedal locomotion stability, but also the real-time object detection, identification and positioning which help the autonomous decision-making and operation during competition. The proposed humanoid robot is equipped with a NVIDIA Jetson TX2 embedded artificial intelligence controller to realize the aforementioned real-time object detection based on the deep convolutional neural network (CNN). Practically, the classes of soccer ball and humanoid robots are detected and recognizes in terms of the You Only Look Once (YOLO) model. The results are validated in the RoboCup 2017, and our team won the 2nd place of the TeenSize league.

A Parallel Robot for Large-range Stiffness Rendering in Three Dimensions

Chao-Chieh Lan

A haptic device is used to transmit impedance to a human user to mimic the impedance of a virtual or real environment. Existing haptic devices use serial or parallel robots to deliver impedance in multiple dimensions. These robots usually have nonconstant Jacobian matrices that result in poor dynamic properties and low impedance stability limits in certain regions within the workspace. To account for these regions, the range of stiffness rendering is limited. This paper presents a three-degrees-of-freedom translational parallel robot with a constant Jacobian matrix in the entire workspace. The consistent dynamic parameters allow a large-range virtual stiffness to be rendered. To provide the accurate and large output force required for high-stiffness rendering, series elastic actuators (SEAs) are used as the input for the parallel robot. SEAs can be used to minimize the geartrain friction and effective inertia to control the output force and impedance more accurately. Design, modeling, and three-dimensional impedance control of the haptic device are presented in this work. Multi-dimensional impedance and virtual-wall control experiments are illustrated to demonstrate the accuracy and rendering range of the haptic device. Since the stable range of virtual stiffness is much larger than existing ones, it is expected that this novel device can be used to render accurate stiffness for both soft and stiff environments.

Safe and Intuitive Manual Guidance of a Robot Manipulator Using Adaptive Admittance Control towards Robot Agility

Daniel Reyes-Uquillas, Tesheng Hsiao

Manual guidance can achieve robot agility effectively, provided that a safe and smooth interaction is guaranteed when the user exerts an external force on the end effector. We approach this by designing an adaptive admittance law that can adjust its parameters to modify the robot compliance in critical areas of the workspace, such as near and on configuration singularities, joint limits, and workspace limits, for a smooth and safe operation. Experimental comparison among various parameter tuning strategies is conducted through a 3D profile tracking test. The results show that the proposed method achieves higher accuracy with shorter task execution time.

Handheld Assistive Robot for Minimally Invasive Spine Surgery

Ping-Lang Yen, Yi-Jing Chu

The hand-held robot has the advantage that the surgeons maintain their original habit and it is flexible for multiple applications by occupying less space in operation room. In this article, starting from the current general cognition of surgical robots, we introduce the composition and technical characteristics of the hand-held surgical robot in view of the needs of minimally invasive spine surgery. Finally, it is proved by experiments that the handheld robot has the ability to assist surgeons to overcome hand tremors and jointly operate surgical instruments to fix the implants along the correct path.

Integration and Application of Sensors and Robotic Technologies

Kuan-Ming Lee, Zong-Hong Lin

Perceptual robots are ideal in real-world environments such as medical and manufacturing sectors through the perceptual activity of human-robot interaction. Difficulties in geometric positioning of sensors due to the structural complexity of perceptual robots and proposing corresponding solutions have been the main challenges in the development of perceptual robots. The integration of 3D electronics into 3D objects through the 3D printing process can be a potential solution for designing actual perceptual robotic systems. The three-dimensional printing process can be divided into the extrusion method and the chemical deposition method. 3D printing offers the advantage of being able to fabricate complex 3D electronic device structures in a single setup, allowing design flexibility and the convenience of customization. Therefore, through the platform of three-dimensional sensor system, and further research on its extension to mechanical, chemical and biomedical fields of perceptual robotics.

Torque Estimation Technique for Feedback Control of Harmonic Drive Actuator

Chun-Yeon Lin, Chih-Yuan Li, Wen-Hao Chen

This paper proposes a torque estimation and control technique for the harmonic drive actuator robotic arm. Torque is estimated from the compliant model of the harmonic drive and applied to the joint torque feedback control. The torque estimation is performed through the relationship between the torsion angle and the torque while measuring the angular position signal by the optical encoder. The torque estimation method is numerically verified and experimentally validated. The proposed method, along with a prototype of the torque sensing system, has been experimentally implemented on the robotic arm for compliance or resistance of the torque and to increase the accuracy.


Inertial Measurement Unit for the Pose Measurement of the Machine Tools

Yan-Xun Lin, Chung-Wei Wu, Yi-Cheng Lin, Wen-Tse Hsiao, Li-Wei Tseng

In this study, the inertial measurement unit (IMU) is installed on the machine tool for the Pose Measurement of the Machine Tools. The acceleration and angular velocity signals measured by the accelerometer and gyroscope of the inertial measurement unit are converted into inclination angle signals. In order to improve the accuracy of measured angle, this study uses the Kallman filter and complementary filter to convert the data of the two signals. Both filtering methods can suppress the noise signals and improve the accuracy.

Quantifying the Glucose Concentration in Urine Test Strip with a Color-calibrated Imaging System

Yu-Hsuan Lin, Cheng-Ru Li, Hsin-Yi Tsai, Chun-Han Chou, Kuo-Cheng Huang

Urine test strips are often used to roughly judge the course of certain diseases, such as kidney disease, diabetes, and hepatobiliary disease. Among them, the development of diabetes can be preliminary assessed through commercial urine glucose test strips. Excessive glucose in the urine will react with the chemicals in the test strip, and then show an appropriate color to provide human observation and judgment. Generally speaking, the color of the test strip can only tell the approximate glucose concentration in urine through visual observation. Accurate quantification must rely on professional electrochemical analysis equipment. However, the system is expensive and has a long detection time. In this study, an imaging system with color correction was developed to quantify the color of urine glucose test paper. Through the composition of high color rendering lighting, darkroom barrel and color algorithm, the measurement results successfully achieved color accuracy with a color deviation of less than 3. A color distribution display method that is more suitable for human observation has been successfully established. The method of swatch contribution analysis has also been developed to numerically quantify the glucose concentration in urine. The results show that the developed system can greatly improve the resolution of urine glucose test strip. This research provides a low-cost, high-quality detection concept that can be used for home medical diagnosis and related chronic disease applications.