24 May 2023 in IEEE Robotics and Automation Letters
This letter proposes a multi-degrees-of-freedom two-wheeled inverted pendulum robot (mD-TWIPR) for off-road transportation. A two-wheeled inverted pendulum robot (TWIPR), popularly known as a segway, has high mobility owing to its simple structure without steering or suspension. The proposed robot has a car-like structure with a low and elongated body that provides advantages for transporting long objects, such as pipes, or transferring patients on off-road terrain. However, a car-like TWIPR structure faces several challenges, including limited acceleration, difficulty traversing slopes, asymmetric payload, and lower ability to climb steps. To address these issues, we implemented a sliding mechanism in the mD-TWIPR. The proposed mD-TWIPR is designed to move without tilting while carrying movable mass, and tilt the torso according to the 20 $^{\circ }$ slope. It can also maintain stationary on a slope with only a small wheel torque by positioning the center of mass on the fulcrum, which is advantageous for transportation operations. Furthermore, it can climb a step using a combination of the wheel and sliding mechanism. The paper discusses the modeling and design specifications and presents a prototype and experiments to verify the proposed performance of mD-TWIPR.
17 October 2025 in IEEE Robotics and Automation Letters
The simulation-to-reality (sim-to-real) transfer of large-scale hydraulic robots presents a significant challenge in robotics because of the inherent slow control response and complex fluid dynamics. The complex dynamics result from the multiple interconnected cylinder structure and the difference in fluid rates of the cylinders. These characteristics complicate detailed simulation for all joints, making it unsuitable for reinforcement learning (RL) applications. In this work, we propose an analytical actuator model driven by hydraulic dynamics to represent the complicated actuators. The model predicts joint torques for all 12 actuators in under 1 $\mu$s, allowing rapid processing in RL environments. We compare our model with neural network-based actuator models and demonstrate the advantages of our model in data-limited scenarios. The locomotion policy trained in RL with our model is deployed on a hydraulic quadruped robot, BeTheX-Q, which is over 300 kg. This work is the first demonstration of a successful transfer of stable and robust command-tracking locomotion with RL on a heavy hydraulic quadruped robot, demonstrating advanced sim-to-real transferability.
13 May 2022 in Journal of Field Robotics
The frequent occurrence of disasters has prompted the development of efficient disaster‐responding equipment. This study deals with the design of special‐purpose machinery and its performance assessment. We defined scenarios through the environmental analysis of disaster situations and expert consulting. We also formulated a set of design specifications by analyzing the objectives of the tasks that must be performed at disaster sites, based on simulated scenarios and existing disaster response machinery. The disaster‐responding special‐purpose machinery was designed to perform various tasks and display rapid movement in disaster situations. Moreover, this paper presents the control structure configured to operate the developed machinery. The performance of the special‐purpose machinery was assessed through different scenario tests.
13 March 2026 in IEEE Robotics and Automation Letters
Designing reward functions for reinforcement learning (RL)-based quadruped locomotion often requires extensive trial-and-error, limiting efficiency and interpretability. Lack of interpretability is particularly critical for large-scale hydraulic quadrupeds, where undetected unstable behaviors during deployment can cause significant mechanical damage. This letter presents a training framework that integrates biologically inspired gait parameters into RL policies, allowing robots to learn locomotion that is both system-aware and human-interpretable. The actor outputs gait parameters—such as gait period, phase offset, stride length, foot clearance, duty factor, and base height—which are coupled with gait-shaping rewards to encourage intent–execution alignment and physically plausible gait patterns. The framework improves training efficiency and provides interpretable signals for monitoring policy behavior. In simulation, we show that locomotion follows the self-generated gait intents as soft constraints, and ablation results demonstrate faster convergence than a pure-RL baseline. We further present reward-term ablation and coefficient sensitivity analyses, indicating that performance is not driven by a single shaping term and is robust to moderate coefficient changes. We validate the approach on BeTheX-Q (over 1.8 m, 350 kg), demonstrating robust real-world locomotion across 2 and 4 km/h walking, stepping-stones, and external disturbances. Finally, gait-parameter analysis reveals interpretable trends that reflect locomotion intent and adaptation across varying conditions.