Jia-Yush Yen

This paper investigates the effects of disturbance observers on a long-travel precision positioning platform. We propose a multi-layer control architecture, including a disturbance observer, a feedforward compensator, gain-scheduling, and control switching. The platform consists of motor and piezoelectric transducer (PZT) stages to enable nanometre-level accuracy within 10 cm travel ranges. We identified the dynamic models of the stages through experiments and applied them to develop control designs. The PZT stage was equipped with feedforward compensators, a disturbance observer and real-time switching control schemes to achieve robust and precise tracking. On the other hand, we applied gain-scheduling and feedforward compensation to the motor stages to track large displacements. The control effects of the integrated platform were validated through simulations and experiments and demonstrated significant improvements in accuracy and robustness. Finally, the platform was incorporated with two-photon polymerisation to fabricate micro-lenses. This work evaluates the lenses’ optical properties to highlight the advantages provided by the multiple control structure for improving precision and microfabrication applications.

In this paper, we propose a real-time model estimation framework using artificial intelligence techniques and apply it to a piezoelectric transducer (PZT) stage equipped with multiple switching controllers. Conventional fixed controllers often fail to satisfy diverse performance requirements: some achieve smooth but slow responses, while others deliver fast yet oscillatory behavior. To address this limitation, we developed a multi-controller switching mechanism that can select optimal control sequences based on predicted system responses, thereby enhancing overall performance. However, the existing mechanism relies on a nominal plant and neglects variations during operation. To address this problem, we employ the eXtreme Gradient Boosting (XGBoost) algorithm to construct a real-time model estimator, which continuously updates the system model during response prediction, thereby improving prediction accuracy. The corresponding controllers are then adjusted according to the updated models and integrated into the switching mechanism to further enhance performance. Finally, we validate the proposed approach through simulations and experiments.

This paper develops multiple control strategies for a precision long-travel stage, which comprises motor and piezoelectric transducer (PZT) stages. First, the PZT stage is equipped with control switching and model estimation mechanisms to achieve nm-level precision within 100 μm distances. The control switching mechanism selects the optimal control sequences by predicting system responses, while the model estimation algorithm updates the system model to improve the prediction accuracy. Second, the motor stage is equipped with gain-scheduling and feedforward control mechanisms to achieve a maximum displacement of 100 mm with a resolution of 0.1 μm. The gain scheduling control modifies the control gain in accordance with tracking errors, while the feedforward control can mitigate phase lags. We integrate the stages to achieve nm-level precision over long travels and conduct simulations and experiments to show the advantages of the control mechanisms. Finally, we apply the long-travel precision stage to fabricate micro-lenses using two-photon polymerization and evaluate the fabricated micro-lenses’ optical characteristics to illustrate the merits of the control strategies.

Objective: Traditional computed tomography (CT)-guided tumor ablation procedures are often hampered by the need for repeated scans and significant radiation exposure, primarily due to challenges in manual needle insertion and the iterative realignment process. This research introduces an advanced intraoperative CT (iCT)-guided robotic system. The core aim is to enhance procedural accuracy and efficiency by implementing precise, real-time remote-center-of-motion (RCM) control for needle interventions. Methods: This study develops an integrated iCT-guided robotic needle biopsy/ablation system. A key distinction from conventional iterative navigation systems is their capability for surgeons to dynamically adjust the intervention line in situ. Real-time visualization of relative positioning is achieved using 3D (three-dimensional) Slicer software. This is coupled with a novel and robust RCM control strategy, meticulously engineered based on screw axis theory and velocity twist, and further optimized with interpolation and feedforward control laws. This framework enables precise, surgeon-guided adjustments of the guiding catheter's posture throughout the procedure. Results: The RCM control system effectively and reliably governed the motion of the robotic guidance arm. This ensured that the catheter tube consistently maintained its accurate orientation relative to the target even after multiple surgeon-adjusted posture modifications. Experimental validation demonstrated the system's high performance, achieving exceptionally low RCM errors as 2×10 ⁻⁴ m. Consequently, any deliberate pause during the robotic arm's movement could be utilized to define a precisely targeted and safe puncture path. Conclusion and significance: Comprehensive numerical modeling and experimental evaluations using a TM5-700 robotic arm with phantoms have provided validation for the proposed screw axis-based RCM control concept. Integrating iCT guidance with this RCM system indicates increased precision, operational flexibility, and overall safety of robotic needle-based interventions. This work presents a noteworthy contribution with potential for clinical adoption in various minimally invasive procedures.

As human–robot interaction becomes more prevalent in industrial and clinical settings, detecting changes in human posture has become increasingly crucial. While recognizing human actions has been extensively studied, the transition between different postures or movements has been largely overlooked. This study explores using two deep-learning methods, the linear Feedforward Neural Network (FNN) and Long Short-Term Memory (LSTM), to detect changes in human posture among three different movements: standing, walking, and sitting. To explore the possibility of rapid posture-change detection upon human intention, the authors introduced transition stages as distinct features for the identification. During the experiment, the subject wore an inertial measurement unit (IMU) on their right leg to measure joint parameters. The measurement data were used to train the two machine learning networks, and their performances were tested. This study also examined the effect of the sampling rates on the LSTM network. The results indicate that both methods achieved high detection accuracies. Still, the LSTM model outperformed the FNN in terms of speed and accuracy, achieving 91% and 95% accuracy for data sampled at 25 Hz and 100 Hz, respectively. Additionally, the network trained for one test subject was able to detect posture changes in other subjects, demonstrating the feasibility of personalized or generalized deep learning models for detecting human intentions. The accuracies for posture transition time and identification at a sampling rate of 100 Hz were 0.17 s and 94.44%, respectively. In summary, this study achieved some good outcomes and laid a crucial foundation for the engineering application of digital twins, exoskeletons, and human intention control.

The gait pattern of exoskeleton control conflicting with the human operator’s (the pilot) intention may cause awkward maneuvering or even injury. Therefore, it has been the focus of many studies to help decide the proper gait operation. However, the timing for the recognization plays a crucial role in the operation. The delayed detection of the pilot’s intent can be equally undesirable to the exoskeleton operation. Instead of recognizing the motion, this study examines the possibility of identifying the transition between gaits to achieve in-time detection. This study used the data from IMU sensors for future mobile applications. Furthermore, we tested using two machine learning networks: a linearfFeedforward neural network and a long short-term memory network. The gait data are from five subjects for training and testing. The study results show that: 1. The network can successfully separate the transition period from the motion periods. 2. The detection of gait change from walking to sitting can be as fast as 0.17 s, which is adequate for future control applications. However, detecting the transition from standing to walking can take as long as 1.2 s. 3. This study also find that the network trained for one person can also detect movement changes for different persons without deteriorating the performance.

Reinforcement Learning (RL) is gaining much research attention because it allows the system to learn from interacting with the environment. Yet, with all these successful applications, the application of RL in direct joint torque control without the help of an underlining dynamic model is not reported in the literature. This study presents a split network structure that enables successful training of RL to learn the direct torque control for trajectory following a six-axis articulated robot without prior knowledge of the dynamic robot model. The training took a very long time to converge. However, we were able to show the successful control of four different trajectories without needing an accurate dynamics model and complex inverse kinematics computation. To show the RL-based control’s effectiveness, we also compare the RL control with the Model Predictive Control (MPC), another popular trajectory control method. Our results show that while the MPC achieves smoother and more accurate control, it does not automatically treat the singularity. In addition, it requires complex inverse dynamics calculations. On the other hand, the RL controller instinctively avoided the violent action around the singularities.

Objective: To investigate the correlation between corneal biomechanical properties and topographic parameters using machine learning networks for automatic severity diagnosis and reference benchmark construction. Methods: This was a retrospective study involving 31 eyes from 31 patients with keratonus. Two clustering approaches were used (i.e., shape-based and feature-based). The shape-based method used a keratoconus benchmark validated for indicating the severity of keratoconus. The feature-based method extracted imperative features for clustering analysis. Results: There were strong correlations between the symmetric modes and the keratoconus severity and between the asymmetric modes and the location of the weak centroid. The Pearson product-moment correlation coefficient (PPMC) between the symmetric mode and normality was 0.92 and between the asymmetric mode and the weak centroid value was 0.75. Conclusion: This study confirmed that there is a relationship between the keratoconus signs obtained from topography and the corneal dynamic behaviour captured by the Corvis ST device. Further studies are required to gather more patient data to establish a more extensive database for validation.

A strain gradient was created by punching a hole in the center of a stretched elastic polydimethylsiloxane membrane to determine the effect of different strains on cultured human keratocytes (HK). In this study, two stretching methods were used: continuous stretching and cyclic stretching. Continuous stretching is relatively static, while acyclic stretching is relatively dynamic. These methods, respectively, represented the effects of high intraocular pressure and rubbing of the eyes on corneal cells. Image processing codes were developed to observe the effects of stress concentration, shear stress, continuous stretching, and cyclic stretching on HKs. The results demonstrate that stretching and shear stress are not conducive to the proliferation of corneal cells and instead cause cell death. A 10% strain had greater inhibitory effects than a 3% strain on cell proliferation. Cell survival rates for continuous stretching (static) were higher than those for cyclic stretching (dynamic). The stretching experiment revealed that cyclic stretching has a greater inhibitory effect on the growth and proliferation of corneal cells than continuous stretching. Accordingly, it shows that cyclic loading is more harmful than high intraocular pressure (static loading) to corneal cells.