論文

車輪型移動ロボットが砂地で構成された斜面を走行する際の制御則について，力学モデルに基づいたフィードフォワード制御，滑りを逐次補償するフィードバック制御をそれぞれ提案した．

© Springer Nature Switzerland AG 2019. This chapter presents a simulator for disaster response robots based on the Choreonoid framework. Two physics engines and a graphics engine were developed and integrated into the framework. One physics engine enables robust contact-force computation among rigid bodies based on volumetric intersection and a relaxed constraint, whereas the other enables accurate and computationally efficient computation of machine–terrain interaction mechanics based on macro and microscopic approaches. The graphics engine allows simulating natural phenomena, such as rain, fire, and smoke, based on a particle system to resemble tough scenarios at disaster sites. In addition, wide-angle vision sensors, such as omnidirectional cameras and LIDAR sensors, can be simulated using multiple rendering screens. Overall, the simulator provides a tool for the efficient and safe development of disaster response robots.

© 2019 IEEE. This paper presents a routing problem of multiple mobile robots collaborating with human workers in a ware-house. In a scenario considered in this paper, the warehouse stores several different of heavy items on multiple stock locations. The human worker pick items up from corresponding stock location, puts them on one of the mobile robot carriers, and then visits another stock location. The multiple mobile robots deployed in the warehouse visit stock locations by following specified routes, carry several items for the worker, and deliver them to a dispatch area in the warehouse. Here, an appropriate number of robots, the maximum speed of each robot, and the load capacity of each robot should be quantitatively analyzed while the routing problem of the multiple robots with the human worker in the warehouse is considered. The research in this paper develops a simulation framework in which the worker and the robots cooperatively performs the pickup and dispatch tasks. An algorithm of the routing problem in the framework is addressed based on the Mixed-integer linear programming, which calculates the order of the picking points for each robot. Simulation studies for an actual warehouse are performed with different conditions between the number of robots, speed, and load capacity. The simulation framework calculates a task time which indicates how long the robots and worker need to complete one shipping list. The task time also quantitatively indicate how large the multiple robots reduce the work load of the human worker.

This paper proposes a prediction model for three-dimensional spatio-temporal soil deformation in bucket excavation. The prediction model for soil deformation (PMSD) consists of two machine learning processes: the long short-term memory (LSTM) and convolutional autoencoder (Conv-AE). These processes use datasets obtained from an experimental apparatus for bucket excavation developed in this work. The apparatus equips multiple depth cameras that precisely capture time-series data of soil deformation in bucket excavation. The LSTM, an extension of a recurrent neural network, successively predicts three-dimensional soil deformation. The Conv-AE is incorporated to both ends of the LSTM in order to quasi-reversibly compress and reconstruct the datasets so that the computational burden of the LSTM is relaxed. Qualitative and quantitative evaluations of the PMSD confirm the feasibility of time-series prediction of three-dimensional soil deformation. The Conv-AE shows sufficient accuracy equivalent to the measurement accuracy of the depth camera. The prediction accuracy of the PMSD is about 10 mm in most of the cases.

Machine learning (ML) plays a crucial role in assessing traversability for autonomous rover operations on deformable terrains but suffers from inevitable prediction errors. Especially for heterogeneous terrains where the geological features vary from place to place, erroneous traversability prediction can become more apparent, increasing the risk of unrecoverable rover's wheel slip and immobilization. In this work, we propose a new path planning algorithm that explicitly accounts for such erroneous prediction. The key idea is the probabilistic fusion of distinctive ML models for terrain type classification and slip prediction into a single distribution. This gives us a multimodal slip distribution accounting for heterogeneous terrains and further allows statistical risk assessment to be applied to derive risk-aware traversing costs for path planning. Extensive simulation experiments have demonstrated that the proposed method is able to generate more feasible paths on heterogeneous terrains compared to existing methods.

A manned pressurized lunar rover is expected to expand the range and the duration of the manned lunar exploration. The dynamic motion of the rover may be different from that on the Earth because such a heavy rover (several tons) travels on loose soil under the low gravity of the moon. Therefore, the main scope of this study is to analyze the dynamic motion of the pressurized rover using a multibody dynamic simulator. In particular, we focus on the steering maneuverability of the rover with regard to vehicle design parameters such as the vehicle tread, wheelbase, and position of the center of gravity. The steering maneuverability of the rover is evaluated based on the wheel slip angle and vehicle rollover stability. The simulation study clarified that the wider tread and lower position of the center of gravity improve the steering maneuverability. Based on the simulation results with different vehicle design parameters and steering input, we summarize allowable vehicle design parameters for improving the steering maneuverability.

Sandy terrain often traps wheeled vehicle or mobile robot with immobilizing wheel stuck. The wheel stuck phenomenon is highly related to wheel slippage and soil failure. Therefore, wheel slip detection and estimation is particularly important for avoiding the wheel stuck phenomenon. This paper proposes a method that can estimate and classify a magnitude of wheel slip using an in-wheel sensor system. The in-wheel sensor captures wheel-terrain interaction characteristics such as contact angles and normal force around the wheel. The proposed method basically estimates a wheel slip by comparing the measured data from the in-wheel sensor with a look-up table generated by a machine learning algorithm. Training data for the machine learning is a variety of experimental data set given from the in-wheel sensor. The look-up table developed in this work distinguishes the magnitude of wheel slippage into three categories: non-stuck wheel, quasi-stuck wheel, and stuck wheel. Experimental demonstration of the proposed method achieves the slip estimation with an accuracy of 90 % or more. Moreover, it is found that tracking the interaction characteristics in a spatiotemporal manner can predict an immobilizing wheel slip or even wheel stuck, resulting in a decrease of mobility hazard.

As part of the robotics technologies required for In-situ resource utilization (ISRU), the development of cargo rovers for transporting resources is needed. However, these cargo rovers have unique technical challenges that differ from conventional exploration rovers, including the need to traverse rough terrains with their varying mass due to transporting payloads. Moreover, research addressing these challenges has been limited, and the relevant technologies have not been fully established. To address these challenges, this paper proposes a parametric model for estimating wheel slippage. The model is formulated as a function of four input parameters: slope angle, rover heading angle, payload mass, and wheel angular velocity, and is applicable to resource-transporting rovers with varying mass. Additionally, the use of a parametric model reduces computational load, which offers advantages for onboard implementation. The proposed estimation model was quantitatively evaluated by comparing datasets obtained from multi-body dynamics analysis. This paper also introduces a new traversability assessment model which incorporates the proposed slip estimation model. We demonstrated the proposed model by integrating it into a sampling based motion planning. The simulation result of the motion planning show that the planner with our model can generate safer motions and enables the rover to reach the target regardless of the cargo payload.

This paper presents slope traversal experiments with slip compensation control for lunar/planetary exploration rovers. On loose soil, wheels of the rover easily slip even when the rover travels with relatively low velocity. Because of the slip, following an arbitrary path on loose soil becomes a difficult task for the rover, and also, the slip will increase when the rover traverses a slope. To cope with the slip issue, the authors previously proposed path following control strategy taking wheel slippages into account. Through numerical simulations in the previous work, it has been confirmed that the proposed control effectively compensates and reduces the slip motions of the rover, and then, the rover can follow a given path. In order to confirm the usefulness of the proposed control for practical application, slope traversal experiments using a four-wheeled rover test bed are addressed in this paper. The control performance of the slip compensation is compared to that of no slip control based on motion traces of the rover in side slope traversal case. Further, the effectiveness of the proposed control is verified by quantitative evaluations of distance and orientation errors. ©2008 IEEE.

Wheeled robots are increasingly utilized for environmental exploration and data collection in environments inaccessible to humans, such as the lunar surface. While adaptive sampling methods like the Upper Confidence Bound (UCB) can balance the trade-off between exploration and measurement data exploitation to model Regions of Interest (ROI), they do not explicitly incorporate robot traversability into the trade-off. Consequently, when faced with multiple scientifically promising locations, the UCB may select paths for the robot that are costly in terms of travel time, leading to time-inefficient surveys, particularly in sloped terrain. To address this limitation, we propose the Slope-aware Upper Confidence Bound (SaUCB), a novel acquisition function that integrates a traversability score based on terrain slope directly into the decision-making process. This allows the robot to explicitly balance exploration, exploitation, and traversal time. Through an extensive simulation study in realistic geological features of lunar terrain models, we demonstrate that the proposed approach generates significantly more time-efficient survey paths. Our method also demonstrates an improved trade-off between investigation time and modeling accuracy within the ROI compared to conventional approaches.

Robotic information gathering methods combining Gaussian Process (GP) regression and Bayesian Optimization (BO) are effective for mapping of soil mechanical properties in uncertain or unknown extraterrestrial terrain. However, standard applications of such methods often model robot traversability as a simple go/no-go constraint, without fully incorporating it as a variable cost that impacts overall efficiency. This limitation is critical on rough terrains like the lunar surface, where traversability heavily influences the trade-off between survey time and modeling accuracy. To address this gap, we propose a comprehensive approach for slope-aware robotic information gathering. The core of this approach is a novel adaptive sampling method, Slope-aware Maximum Uncertainty Sampling (SaMUS), which features a new BO acquisition function that explicitly incorporates terrain slope to balance uncertainty reduction with path traversability. We demonstrate the effectiveness of our method through an extensive simulation study in two environment types: controlled basic terrains for qualitative analysis and realistic lunar terrains for evaluating the performance of the proposed method. The results show that the proposed approach generates significantly more time-efficient survey paths by strategically avoiding highslope regions. Quantitative analysis on the realistic terrains further demonstrates that our method achieves a better tradeoff between survey time and modeling accuracy, particularly on flat and steep terrains. Finally, we provide practical guidelines for selecting optimal parameters based on terrain class and operational time constraints.

Grip, walk, crawl, and jump. Soft robots are integrated functional structures composed of compliant mechanisms, whose activity spans various industrial applications such as surgery, healthcare, surveillance, and even planetary exploration. One of their promising mobility mechanism is snap-buckling; the instability mode of flexible structures passing from one equilibrium state to another can instantaneously generate large power for its motion. Predicting their performance with even simple geometry requires disentangling material, geometric nonlinearity, and contact, thereby still being a challenging problem to date. Here, we study the jumping dynamics of hemispherical elastic shells driven by snap-buckling, as a model system of soft jumping mechanisms, combining experiments, simulations, and analytical theory. We find that the contact transition dynamics trigger the jumping phenomenon upon snap-buckling by constructing the analytical predictions with shell elasticity in excellent agreement with both experiments and simulations. Despite the simple geometry of the shell, its dynamical performance primarily relies on a complex interplay between elasticity, geometry, and contact friction. By elucidating the dynamics of the building blocks of soft robots that undergo large deformations, we can build their predictive experimental and numerical framework. Our research paves the way for designing soft robots suitable for the required loading conditions or structural requirements without empirical methods.

This paper presents a perception-aware path planning framework for unmanned aerial vehicles (UAVs) that explicitly considers perception quality of a light detection and ranging (LiDAR) sensor. The perception quality is quantified based on how scattered feature points are in LiDAR-based simultaneous localization and mapping, which can improve the accuracy of pose estimation of UAVs. In the planning step of a UAV, the proposed framework selects the best path based on the perception quality from a library of candidate paths generated by the rapidly-exploring random trees algorithm. Consequently, the UAV can autonomously fly to a destination in a receding horizon manner. Several simulation trials of the photorealistic environments confirm that our proposed path planner reduces pose estimation error by approximately 85 % on average as compared with a purely-reactive path planner.

In future lunar/planetary exploration missions, wheeled rovers will be promising robots to perform wide-area explorations. Wheeled robots, however, are prone to slip in soft and uneven terrain, and therefore, securing the mobility performance is a critical problem. This paper first describes a simple soil-wheel interaction model which enables slipratio estimation from a combination of wheel load and traction. Second, a drive control method for maintaining vehicle mobility is discussed and a six-wheeled robot which implements the control system was developed. The robot has six wheels and six vertical suspensions with built-in linear actuators. Load cells are installed between each wheel and suspension to monitor the wheel load, and the positions of wheels are individually controlled by moving the suspensions up and/or down so as to keep the desired wheel loads. In this study, a simple algorithm for distributing equivalent load to every wheel was implemented in the control system. Experiments were performed and the effectiveness of the drive control method was examined.