The Koopman operator framework has shown promising results in enabling the analysis of nonlinear dynamics into an infinite-dimensional linear representation. Koopman direct encoding (KDE) is a model-based approach that utilizes inner products and compositions in a Hilbert space to compute the Koopman operator. However, it has primarily been applied to autonomous systems and simulation environments. Here, we extend the application of KDE to nonautonomous systems and real-world environments by introducing Koopman direct encoding-based model predictive control (KDE-MPC). It was validated on nonlinear electromechanical systems with segmented dynamic conditions, such as contact-noncontact transitions, which pose challenges for modeling and control. Simulation results demonstrate a more stable and smoother position profile compared to proportional-integral-derivative control, particularly at discontinuous boundaries. KDE-MPC was also applied to real-world systems, achieving similar position tracking performance to simulation results. We anticipate that KDE-MPC will offer a viable solution for complex robotic control challenges.

When manipulating an object, a robot must recognize not only the parts it directly grasps but also the surfaces in contact with the environment, which we refer to as extrinsic contact surfaces. These surfaces directly affect how the object interacts with its environment, and accurate surface estimation is critical for precise robotic manipulation. This study presents a novel framework for extrinsic contact surface reconstruction using vision-based tactile sensing. By leveraging marker-based tracking and analyzing kinematic constraints, we classify contact types and estimate the locations of both point and line contacts. To reconstruct the extrinsic contact surface, we compare three data integration methods: Mixed Vector Approach (MVA), Orthogonal Distance Regression (ODR), and Random Sample Consensus (RANSAC). Experimental results demonstrate that MVA achieves the highest accuracy in most cases by effectively integrating contact data while minimizing randomness. Experiments conducted on various object geometries validated the robustness of the proposed method, achieving an average positional error of 4.15 mm and an angular deviation of 4.58°. The results confirm that extrinsic contact sensing enables more efficient and precise object shape estimation, providing a promising approach for robotic manipulation.

Humans perform grasping by breaking down the task into a series of action phases, where the transitions between the action phases are based on the comparison between the predicted tactile events and the actual tactile events. The dependency on tactile sensation in grasping allows humans to grasp objects without the need to locate the object precisely, which is a feature desirable in robot grasping to successfully grasp objects when there are uncertainties in localizing the target object. In this paper, we propose a method of implementing a tactile event based grasping algorithm using memorized predicted tactile events as state transition triggers, inspired by the human grasping. First, a simulated robotic manipulator mounted with pressure and vibration sensors on each finger, analogous to the different mechanoreceptors in humans, performed ideal grasping tasks, from which the tactile signals between consecutive states were extracted. The extracted tactile signals were processed and stored as predicted tactile events. Secondly, a grasping algorithm composed of eight discrete states, Reach, Re-Reach, Load, Lift, Hold, Avoid, Place, and Unload was built. The transition between consecutive states is triggered when the actual tactile events match the predicted tactile events, otherwise, triggering the corrective actions. Our algorithm was implemented on an actual robot, equipped with capacitive and piezoelectric transducers on the fingertips. Lastly, grasping experiments were conducted, where the target objects were deliberately misplaced from their expected positions, to investigate the robustness of the tactile event based grasping algorithm to object localization errors.

Despite the advantages of electromyography (EMG), which can grasp intent before actual movement, there have not been many studies on the use of EMG in exoskeletons. In this paper, we conducted an experiment to analyze the characteristics of electromyography when EMG signals are used in exoskeleton robots. We integrated the advantages of EMG sensors and physical sensors to control exoskeleton robots using EMG signals and physical signals. The walking environment was defined using the surface electromyography (sEMG) signal and the physical signal to an accuracy of 88% or more. To compensate for the limitations of physical sensors, we used sEMG to distinguish changes in the load during walking. Moreover, sEMG signals and physical signals were used to distinguish external collisions and to identify other variables that could be distinguished. Finally, to examine the characteristics of muscular fatigue, which is a disadvantage when using electromyography, we conducted a muscle fatigue experiment in the lower limb and summarized how the EMG characteristics change in relation to the degree of force and the muscle fatigue.