The development of bionic hands is a crucial area robotics research, aiming to achieve human-like dexterity, adaptability, and efficiency in manipulation tasks. This study presents the Skeleton Bionic Hand (SkB-Hand), which features a dual elastic-tendon mechanism for extensor and volar plate in a single finger. The skeleton structure ensures the SkB-Hand is lightweight (< 600g) and low-cost (< 150USD) while maintaining performance in various grasping tasks. We evaluated the SkB-Hand through experiments such as the Kapandji test, GRASP Taxonomy, and dynamic scenarios. Results showed the SkB-Hand scored 7/10 in the Kapandji test and 33/33 in the GRASP Taxonomy test. The hand demonstrated resistance to deformation under external forces while ensuring flexibility. Integrated into the Techman Robot (TM-Robot), the SkB-Hand provide its potential for general robotic grasping tasks in daily life.

Most existing P300 brain computer interface (BCI) based brain-controlled wheelchairs (BCWs) position visual stimulus panels in a manner that requires users to frequently shift their gaze between the control panel and the surrounding environment. This constant gaze shifting can reduce intuitiveness and compromise safety. To address this issue, this study introduces a translucent visual stimuli control panel placed directly within the user's field of vision. Visual stimuli, representing movement commands (forward, backward, left, right, and stop), are projected onto this panel using a micro-projector positioned 35 cm away from the user. Notably, the location of these flashing stimuli dynamically aligns with the user’s eye gaze, which is detected through electrooculographic (EOG) signals. Consequently, the control interface naturally follows the user’s visual attention, enhancing both intuitiveness and operational safety. Classification of the P300 signals for wheelchair control is performed using a support vector machine (SVM), while the canonical correlation analysis spatial filter enhances the information transfer rate. Performance assessments involved 10 healthy subjects navigating wheelchair trajectories "U" (5.7 m) and "S" (12.4 m) shapes. Experimental results confirmed that the eye-gaze-guided BCI approach significantly improves accuracy and user experience compared to traditional non-EOG-guided methods.

This study introduces a novel frequency-based attention network (FANet) designed to acquire frequency information in surface electromyographic signals (sEMG) while mitigating potential high-frequency disturbances. The utilization of the discrete cosine transform (DCT) in dedicated attention modules, enables the conversion of sEMG features to the frequency domain and facilitating the extraction of spectral characteristics. This attention module replaces the self-attention module within the multilayer perceptron (MLP) based transformer, offering a streamlined means for internal frequency features analysis in deep learning models. In evaluation, we focus on the dominance of the “rest” gesture in available NinaPro DB4 and NinaPro DB5 datasets. Therefore we employ common metrics tailored to the class-imbalanced hand gesture classification task. Through comprehensive experiments and ablation studies, the results showed significant improvement in hand gesture recognition, with the DCT-based attention modules achieve accuracies of 78.70% and 89.64% in the NinaPro DB4 and DB5 datasets, respectively, while maintaining low computational cost with log-linear complexity.

Mechanical engineering (ME) students have specific domain knowledge in developing physical systems, such as machines, vehicles and production equipment. Nevertheless, the strong demands of utilizing artificial intelligence (AI) for physical systems to become a cyber-physical system (CPS) is now a commonplace of modern physical systems. Hence, it is necessary to cultivate greater AI design and programming capabilities in ME students. In this poster, the "Intelligent Systems: Theory and Practice" course (ME5061 in ME, National Taiwan University) for senior undergraduate/ graduate students) is presented. This course has been offered once in the spring semester of 2022, and 34 students participated in this course.

This poster aims at providing an interdisciplinary course (Medical Mechatronics and Control/ ME5053 in National Taiwan University/ Senior Undergraduate/ Graduate Students) designed to encourage the mechanical engineering (ME) students to learn about machine learning (ML) through project-based learning (PBL). Our syllabus design involves the electromyography (EMG) data collection, datasets and ML, and robotic manipulator control through recognizing EMG patterns via ML. Thus, students in ME become more attracted to learning computer science ML techniques through participation in robot-related courses via recognizing and classifying real EMG data.

While the ability to perceive attitude in 3D trajectory of the movement is a basic concept of a motion capture, a 9-axis inertial measurement unit (IMU) device containing accelerometer, gyroscope and magnetometer is introduced as a challenge for motion tracking device. In this paper, a modular architecture in terms of accuracy on orientation of the sensor output, implemented for wearable motion capture system in attitude estimation. It contains 13 IMU module devices as a sensory network attached in the human body for reconstructing human attitude estimation. The system is integrated as quaternion-based implementation where several methods are separately used for achieving better result. The first proposed methods are using digital motion processor for gaining accurate data and directly calculated inside the sensor itself. Secondly, a complementary filter (CF), fused using gradient descent algorithm. Another approach method is combining CF with Mahony filter. Furthermore, all of approaches are working with DMP system to minimize the accumulative errors. As a validation, a passive manipulator robot containing encoders is used for comparison. The experimental results showed that all of the proposed methods can work properly and they represented acceptable performances and achieve accuracy for orientation.

This paper proposes a reconfigurable biopotential amplifier for a medical instrumentation course in the department of electrical engineering. With the reconfigurable biopotential amplifier, the students are able to measure their biopotential signals such as electromyography, electrocardiography and electrooculography in the class. In addition, reconfigurations of the circuit board can be achieved by forming different combinations of the circuit modules to investigate the effects of employing the amplifiers and filters, as well as to practice the topics of various biopotential signal sources in terms of adjusting the configurations of the circuit board. At the same time, variable resistors are further used to adjust the amplifier gains and frequency conditions. Moreover, a microcontroller is provided for converting the analog signals to digital forms so that the biopotential signals can be displayed and recorded with a computer program. Therefore, the reconfigurable biopotential amplifier is a low cost and compact educational tool. The students can practice this module only with a personal computer which connects a USB node to provide the DC power and serial data communication. In addition to the improvement on the learning achievements for the biopotential signal measurement experiments within the lecture topics in the medical instrumentation course, the reconfigurable biopotential amplifier also benefits the hands-on skills of students for designing the amplifiers and filters in medical instrumentations.

This paper presents a wireless sensor network approach for collecting rehabilitation data with exercising bilateral sanding units. Three wireless sensor nodes are attached to a conventional sanding unit for automatic data collections. Two wireless sensor nodes are desired for a set of sanding hand blocks, and the remaining one is used for the reciprocal exercise platform. Three wireless sensor nodes can be easily attached to conventional sanding units with less wiring and setup efforts for fast realizing data collection functions. Finally, a system prototype is produced in this project, and the bilateral accelerations, exercise strokes and gripping forces are also properly collected to validate our approaches.

In this paper, the wireless sensor node technology is used to design a pushrim activated wheelchair (PAW) controller. The proposed wireless sensor node is capable of collecting user's toques which are applied on two pushrims to generate wheelchair speed commands. In order to control a PAW smoothly, parametric sigmoid functions are developed to accelerate or decelerate wheel motors based on users' setups. In addition, the state transitions among forward, backward, and stop of wheel motors are also well arranged. Three wireless motes are used in this system, where two of them are desired for wheel motor controllers, and the remaining one serves as a network coordinator. A supervisory control is further developed to adjust the resistance, operation time factor and left-and-right bias to meet different wheelchair manipulation purposes. Finally, a wheelchair prototype is developed based on the proposed architecture to verify our approaches.

This paper describes an electrooculography (EOG) based human-wheelchair interface for wheelchair users. A pair of electrodes which measures the eye-gaze direction of users is desired as wheelchair manipulation commands. In addition to EOG based wheelchair manipulations, this paper also introduces ultrasonic arrays for detecting distances to actively avoid collisions. In order to simplify setup procedures of using this system, two electrodes are mounted on two side arms of eyeglasses respectively that are possible to have tight contacts with skins around eyes. Therefore, wheelchair movements are capable of following eye-gaze directions. On the other hand, an analog biopotential signal amplifier and a laptop computer are used to develop the proposed EOG based human-wheelchair interface controller. Finally, experimental results demonstrate the operations of EOG based human-wheelchair interface.

Powered wheelchairs are popular mobility aids for most disabled patients and elders. Most of powered wheelchairs are manipulated via a conventional joystick. The powered wheelchair user has to control the direction and velocity of a powered wheelchair in a very narrow joystick operation range. Therefore, the user's joystick command may be very sensitive to the wheelchair motions. In addition, such an operation is not analog to conventional manual wheelchairs. In this paper, a power assisted wheelchair control system is developed to perform similar operations of manual wheelchairs, while the efforts of pushing wheel rings are significantly reduced for the flat and barrier free ramp terrains. For the proposed power assisted control system, electric motors are driven according to the tactile forces between user's hands and wheel rings. The tactile forces are further combined with slopes of ramps and resistances of operations to generate wheel velocities in terms of an intelligent fuzzy control system. Therefore, the proposed power assisted wheelchair not only maintains user's motor function of upper limbs but also performs easy driving for a long distance. The performance is evaluated practically using an electromyogram (EMG) instrument that is used to measure the muscle activities of volunteers on their upper limbs. Experiment results show that the power assisted function significantly reduces the pushing or pulling forces under flat and barrier free ramp terrains.

Fixed partial denture produced by conventional method is greatly relied on the skill and experience of dental technician. Its quality and accuracy depends mostly on the technician's subjective judgment. In addition, the manual process involves many procedures which require a long time to complete. Most important, it does not preserve any quantitative information for future retrieval. In this paper, a new device for scanning denture images and reconstructing 3D digital information of dental model by abrasive computer tomography (ACT) was designed in-house and proposed. The fixed partial denture was then produced by the rapid prototyping (RP) and computer numerical control (CNC) machining methods based on ACT scanned digital information. A force feedback sculptor (freeform system, sensible technologies Inc., USA) accompany with 3D Touch technology was applied to modify the morphology or the design of denture. It enables the dentist to perform digital manipulation of denture profile with real-time and user-friendly human interactive operation without relying on CAD/CAM technician or dental technician. In this article, the comparison of conventional manual operation and digital manufacture using both RP and CNC machining technologies for denture production was presented. In addition, a digital custom denture manufacturing protocol integrating proposed computer abrasive teeth profile scanning, computer-aided denture design, 3D touchable force feedback feature modification and advanced denture manufacturing techniques were suggested. These proposed methods provide a solid evidence that digital design and manufacturing technologies may become a new avenue for custom-made denture design, analysis, and production in 21st century.

Autonomous robotic wheelchairs are widely discussed in recent years. Most of autonomous robotic wheelchairs were developed based on rich computation architecture to deal with complicated navigation efforts such as personal computers or embedded computers. However, such rich computation architectures increase cost of robotic wheelchair when compared to conventional powered wheelchairs. At the same time, large amounts of energy consumptions, lower reliability, and inefficient size of personal computers reduce the practicability of robotic wheelchair. In this paper, we present a low cost solution to solve complicated wheelchair navigation problem based on distributed computing architecture. This solution is developed based on the programmable System-on-Chip (PSoC) computing and control architecture. The tasks within autonomous navigations are categorized into human machine interface, sensor collection, fuzzy logic based navigation functions, closed loop motion controller, and these tasks are individually implemented using the PSoC. Finally, the task oriented PSoC are integrated and fused to perform autonomous navigations of robotic wheelchairs. Consequently, the proposed PSoC based distributed computing architecture was successfully implemented and experimentally tested. The navigation results were verified using the results inferred from the personal computer based computing architecture.

Coordination of multiple autonomous robots is a typical research problem in multi-agent systems. The distributed agent models that represent the supervisory controllers of autonomous soccer robots are capable of communication, interaction and collaboration. By introducing the multi-agent based control architecture, the soccer robots perform seamless team cooperation to achieve the desired mission during competition. Due to the multi-soccer robot system is a discrete event dynamic system with concurrent and asynchronous characteristics, the distributed agent-oriented Petri net (DAOPN) is selected in this paper to model and control the behaviors of the multi-soccer robot systems. The agent models proposed in this work consist of the coordination model and the skill oriented robot behavior models. Especially, the agent based communication places enable the distributed agent models to be communicated directly. In this manner, the DAOPN based agent models can act as the robot controllers and coordinator to perform run-time controls of the soccer robot team. Therefore, the proposed approach not only resolves problems of the multi-robot coordination, but also provides a feasible way to overcome the complicated mathematical functions or rule bases. Finally, this system is tested using four soccer robots, and the results verify the proposed DAOPN based controller.

Rehabilitation plays an important role in the future aging society. However, due to increasing aging population, the large amount of therapy efforts and spaces for rehabilitation may be not sufficient. Therefore, the tele-rehabilitation approach is proposed to resolve the shortages of physical therapists and rehabilitation spaces. The tele-rehabilitation is a technology that monitors and controls the rehabilitation equipments in efficient and convenient manners through the network to achieve remote rehabilitations. To achieve efficient and safe tele-rehabilitations, a real-time rehabilitation platform for database accessing, message exchange, signal and status monitoring, and remote manipulations of rehabilitation equipments are desired in this paper based on distributed object-oriented techniques. The proposed distributed object-oriented tele-rehabilitation platform also achieves the purposes of robustness and loading balance issues. In addition, a service allocation algorithm is proposed to dynamically balance the overall loading of each host server so that the service quality of the platform can be improved. The performance of the CPU and utilization of the dynamic memory of the host servers are all measured and recorded periodically as the bases of allocating new services and re justifying executing services in real-time. Finally, the fault detection and recovery algorithm is proposed to construct robust tele-rehabilitation platform. Therefore, the proposed tele-rehabilitation platform achieves time-efficient and cost-effective purposes.

This paper presents an active vision based space-positioning robot (AVBSPR) for surgical navigation applications. The development of the proposed navigation robot consists of a five-axis SCROBOT-ER VII Robot, a stereoscopic based eye-in-hand image capture module and a set of infrared (IR) based surgical marker emitters. To increase the navigation space and to determine the best image capture orientation, a fuzzy inference engine is constructed to dynamically track the moving surgical markers. On the other hand, the stereoscopic image coordinates are used to calculate the 3D space coordinates of the surgical markers for surgical navigation applications. Finally, a prototype system had been implemented, and several experimental tests had been done. The experimental results showed that the robot tracks the surgical markers well, and the positioning error is around 5 mm within 2000 mm distance operating ranges.

The surface mounting technology (SMT) assembly procedures consist of calibrating the printed circuit board (PCB); vacuuming components form the feeder stations; compensating the orientation of the vacuumed surface mountable component (SMC); placing SMC chips on PCB. To increase the throughput of the assemblies, the synchronous vacuuming of SMC components is designed. Evaluating the cycle times of SMT assemblies requires the modeling of the specified times of image processing, vacuuming and releasing components and the route dependent robot movement times. In this paper, the route oriented Petri nets (ROPN) is newly proposed to solve the SMT assembly problems in flexible routes. Based on the ROPN modeling capabilities, the flexible component mounting routes and the synchronous vacuuming procedures can be modeled actually. Finally, two actual testing PCB board data were evaluated and the results had been compared with the actual cycle times to verify the proposed approaches.

In a 300 mm IC foundry fab, the intrabay automation systems are more important than a 200 mm one. In general, the intrabay automation systems include the automated controls of the process tools, overhead hoist transports (OHTs) stockers, and rail shortcuts. These control systems are operated concurrently. They must be cooperated seamlessly to increase the system performance and to avoid deadlocks. In this paper, the behaviors of the intrabay automation systems are modeled using the distributed agent oriented Petri net (DAOPN). The fab intrabay DAOPN models are categorized into the new order generation, OHT vehicle dispatching, zone control, path decision, and tool processing models. Finally, the DAOPN is defined, and the models of the intrabay system with one stocker, five tool stations, one OHT loop, and two rail shortcuts are modeled and simulated using the proposed DAOPN. The simulation results are evaluated based upon the conditions of changing the number of OHT vehicles and the dispatching rules.

In this paper, we practice an Internet-based in-house healthcare and home automation system. In this work, the components of data acquisition, voice dialing, video grabbing, status monitoring, and Web service are all designed and implemented. Especially, the client-server architecture is used to integrate the hardware and software components. Therefore, the proposed system can be used in the rapidly aging society. The family and the social service related department or agencies are able to take more notices to the elder person who is living alone via the network. The prototype of this system had been implemented. Finally, a demonstration board that simulates the operations of security alarming, appliance control, and data acquisition are also developed to test the Internet-based in-house healthcare and home automation system.

The fast growing computer technology had been successfully applied to the industry and medical applications. The operation efficiency, system accuracy, and quality consistency can be significantly improved by introducing the computer techniques. In recent years, several works combine the techniques of the computer software, machine vision and image processing to assist the surgeons to improve the surgical quality. Especially, the surgery navigation system is more emphasized than ever. In this paper, we propose a computer aided retained auricular prosthesis surgery system. In this work, a 3D digitizer is used to measure the coordinate and posture of the patient. The patient's coordinate is further transformed to the coordinate of the computer model. In addition, a sound-based navigation system is developed to guide the surgeon to reach the registration position that was identified at the pre-surgery stage. Finally, a clinical case of the retained auricular prosthesis surgery of the microtia children was proposed and successfully validated.

Integrated circuit (IC) foundry fabs are difficult to model due to the increasing product mixes and flexible routes. An imprecise fab model cannot be used to evaluate fab performance or estimate product cycle time. The distributed colored timed Petri net (DCTPN) is a type of high-level Petri net with embedded entity attribute, time, communication, and remote object invocation properties that contribute to realistic descriptions, distributed modeling, and manufacturing execution system integration. Furthermore, the resource-oriented DCTPN modeling extends product mix and flexible route modeling capabilities. The DCTPN conflict resolution and token competition rules are used to construct the dispatching system for a controlled IC fab. A simplified 200 mm IC fab with six functional areas is discussed and demonstrated based on different control policies.

In a 300 mm integrated circuit (IC) foundry fab, the intrabay automation is more important than a 200 mm one. In general, the intrabay automation systems include the automated controls of the process tools, overhead hoist transports (OHTs), stockers, and rail shortcuts. These control systems are operated concurrently. They must cooperate seamlessly to increase the system performance and to avoid deadlocks. In the paper, the behaviors of the intrabay automation systems are modeled using the distributed agent oriented Petri net (DAOPN). The fab intrabay DAOPN models are categorized into the new-order generation, OHT vehicle dispatching, zone control, path decision, and tool processing agent models. Finally, the DAOPN is defined, and the modeling procedures of the intrabay behaviors are all illustrated. Consequently, an experimental intrabay OHT loop is discussed and evaluated using the proposed DAOPN models and the DAOPN simulator.

The reconfigurable manufacturing system (RMS) is a new concept, and it can increase the product variety and production quantity as well. The key factor of introducing the RMS depends on the redesigned equipment layout and material flows. In order to construct a high performance production system, the allocation of resources must be analyzed first. The paper proposes an equipment layout assignment algorithm (ELAA) to find the near optimal equipment layouts for reducing the cycle time of core products. Consequently, the paper uses the distributed colored timed Petri net (DCTPN) to model the RMS, and the new configuration can be designed based on the resource allocation results. In addition, the performance of the reconfigured manufacturing system is also evaluated based on the DCTPN simulation results. Finally, a simplified 200 mm fab is discussed to demonstrate the procedures of constructing reconfigurable fab models, allocating resources and evaluating the performance.

Proposes the modeling and performance evaluation of a controlled IC foundry fab using a distributed colored timed Petri net (DCTPN). A DCTPN is designed for highly model-mixed and flexible routing manufacturing systems. The distributed models can be integrated through communication places, and the manufacturing execution system (MES) can be executed using a COM (component object model) server place. Especially the process activities of a general IC foundry tool are analyzed, modeled and grouped into a basic tool model library. Based on the tool model library, an entire fab model can be constructed. The DCTPN-based fab model can act as a virtual fab, and be used to estimate fab performance and fab behaviors. In addition, the DCTPN conflict resolution and token competition rules are used to control a fab. Finally, a simplified 200 mm IC fab is presented and its system performance, including throughput, stage move, WIP (wafer in process) distribution, lot cycle time, and utilization, is evaluated. The operation histories of tools and lots are also included. The entire fab model has been verified in a real IC foundry fab.

The performance of a flexible manufacturing system (FMS) depends on the equipment efficiency and process control. In order to increase the equipment efficiency, the failure mode analysis and fault diagnosis can be used to reduce the frequency of unexpected breakdowns of machines. In addition, the statistical process control (SPC) can be used for adjusting the process parameters to eliminate process variations. In this paper, a colored timed Petri net (CTPN) is used to model the process behavior of an FMS. In addition, the CTPN-based SPC, fault diagnosis, and failure model and effect analysis are modeled and analyzed individually. Especially, all of the modular models are integrated and linked based on the CTPN. Due to the unified CTPN modeling, the information of each modular model in the entire system can be exchanged and integrated directly and efficiently. Finally, the entire CTPN FMS models are implemented using G2 real-time expert system. Consequently, the proposed CTPN-based simulator can be acted as a real-time FMS monitor and controller through the G2 standard interface.