Published papers

The bucket drum, a cylindrical rotating mechanism that efficiently excavates and collects lunar regolith, is essential for in-situ resource utilization (ISRU). While bucket drums offer advantages in energy efficiency and adaptability to low-gravity conditions, optimizing their shape and motion remains an open issue. Therefore, this paper aims to find an optimal design of the bucket drum using the discrete element method (DEM). We focus on five key design parameters were analyzed: two related to bucket shape (scoop throat length and scoop inlet number) and three related to motion (bucket vertical force, horizontal velocity, and angular velocity). These parameters were evaluated based on two performance indices: sand fill ratio and excavation power consumption. By solving the multi-objective problem of maximizing fill ratio while minimizing power consumption, an optimal design of the bucket drum was identified. Subsequently, a bucket drum with the optimal shape was 3D-printed, and the optimal motion of the bucket drum was applied using the soil excavation test apparatus. The test results qualitatively matched the results derived from the DEM analysis. This highlights the validity of the relationship between the design parameters and two performance indices, contributing to the advancement of ISRU excavation technologies.

In this paper, we propose an action planning algorithm and its evaluation method based on dynamic simulation for a novel type of hybrid leg-wheel rover for planetary exploration. Hybrid leg-wheel robots are recently receiving a growing interest from the space community to explore planets, since they offer an appropriate solution to gain improved speed and mobility on unstructured terrain. However, in order to fully reach the hybrid mechanism's potential, it is necessary to establish an optimal way to define when to use one over the other locomotion mode, depending on the soil conditions and topology. Even though this step is crucial, little attention has been devoted to this topic by the robotic community. The switching of motion mode, that is either wheel or leg are the actions to be planned, that we are considering in this paper. We aim at generating the safest and the least energy demanding path to reach a point of scientific interest. In order to define the optimal path with the set of switching actions required for the robot to follow it, the authors developed an action planning algorithm and a path evaluation method based on a four steps approach. First, an optimal candidate path on a rough terrain is generated based on topology and specifications' criteria functions. Then switching actions are defined along this path depending on the hybrid robot's performances in each motion mode. The next step is a dynamic simulation of the robot controlled to follow the path. Finally, the path is evaluated based on the energy profile spent by the actuators and calculated by the simulation. Demonstrations for the proposed technique are addressed along with a discussion on characteristics of the candidate path and the energy profile of the robot. ©2008 IEEE.

Estimating the position of a robot is an essential requirement for autonomous mobile robots. Visual Odometry is a promising localization method in slippery natural terrain, which drastically degrades the accuracy of Wheel Odometry, while relying neither on other infrastructure nor any prior knowledge. Visual Odometry, however, suffers from the instability of feature extraction from the untextured natural terrain. To date, a number of feature detectors have been proposed for stable feature detection. This paper compares commonly used detectors in terms of robustness, localization accuracy and computational efficiency, and points out their trade-off problems among those criteria. To solve the problem, a hybrid algorithm is proposed which dynamically switches between multiple detectors according to the texture of terrain. Validity of the algorithm is proved by the simulation using dataset at volcanic areas in Japan. © 2013 Springer-Verlag.

Lunar infrastructure construction involves leveling of the lunar surface and collecting lunar regolith as a building material. RASSOR 2.0 developed by NASA is one of the typical robotic vehicles as a lunar excavator. It features cylindrical rotating bucket drums for collecting regolith, positioning it as a pivotal tool for future lunar in-situ resource utilization and infrastructure development. However, the optimization of the bucket drum's shape and its motion remain as open issue. Therefore, this study aims to find an optimal design of the bucket drum through numerical simulation using the discrete element method (DEM). We introduce five key design parameters of the bucket drum: two of them are related to the bucket shape (scoop throat length, scoop inlet number) and the rest of them are to motion (bucket vertical force, horizontal velocity, and angular velocity). These five parameters were examined in accordance with two performance indices: the sand fill ratio in the drum and the power consumption of the excavation. Solving the multi-objective problem of increasing the fill ratio and reducing power consumption, we found an optimal bucket drum shape and motion. Subsequently, a bucket drum reflecting the optimal shape was fabricated, and the optimal motion was tested. The test results qualitatively matched with the results derived from the DEM analysis. This outcome highlights the validity of the relationship between the design parameters and two performance indices.

Offroad mobility in extraterrestrial exploration can be significantly degraded by vehicle slippage on soft terrain covered with fine powdery sand known as regolith. Comprehensive design analysis and verification of the rover's mobility have been widely investigated using numerical and experimental approaches. Numerical simulations for the mobility analysis require accurate modeling of the contact forces between the wheel and sand. While several wheel-sand interaction models have been proposed based on the Bekker-Wong-Reece theory, the Dynamic Resistive Force Theory (DRFT) proposed in the early 2020s is a notable approach for its low computational cost and adaptability to quasi-dynamic motion. The scaling factor used in DRFT is the only parameter for representing sand-dependent parameters, and its value needs to be empirically tuned to achieve accurate and reliable mobility analysis. Therefore, the authors have integrated a single-wheel test bed into a closed-loop Hardware-In-the-Loop Simulation (HILS) associated with DRFT. This HILS-DRFT experimentally observes the characteristics of the wheel sinking phenomenon in the wheel test bed, and then, the value of the scaling factor is tuned in real-time by the difference between the observed wheel sinkage and the DRFT simulation. This real-time parameter tuning method accurately reproduces wheel mobility, particularly in its transient states, which were previously unachievable with DRFT alone. The proposed HILS-DRFT with the real-time tuning method will contribute to the efficient and reliable development of rovers.