Advanced Robotics

Uemura, Mitsunori; Tsujikawa, Shota; Suefuji, Masayoshi; Arita, Hikaru

doi: 10.1080/01691864.2024.2420094pmid: N/A

To operate in uncertain environments, robots must dynamically interact with and recognize objects. This paper proposes a novel control method to dynamically estimate the center-of-mass of an uncertain object. In this method, a robot finger moves an object, and the moving direction of the finger is dynamically adjusted to estimate the direction of the object's center-of-mass (friction center). The primary advantage of this method is its capacity to rapidly estimate the center-of-mass direction utilizing a single contact. Once the direction of the center-of-mass is determined, this information can be used, for example, for the robot hand to grasp the object securely by surrounding its center-of-mass, ensuring stable handling without unexpected movements. Moreover, determining the center-of-mass can aid in automatically producing training data for machine learning applications. Both simulation and experimental results validate the proposed control method's efficacy, demonstrating its ability to quickly converge to the desired state. This control problem poses a significant challenge as the system has fewer actuators than degrees of freedom, indicating that this controlled system is underactuated. Nonetheless, the proposed control method operated successfully.

Hu, Site; Horii, Takato; Nagai, Takayuki

doi: 10.1080/01691864.2024.2415995pmid: N/A

The growing ubiquity of autonomous robots across different fields necessitates that agents adapt to diverse tasks and ensure transparency and intelligibility in their decision-making processes. This study presents a novel framework that combines a graph-structured world model with large language models (LLMs) to address these requirements. First, a latent space is created to capture the reachability between distinct states. Next, a graph is constructed within this latent space by clustering an offline dataset that effectively captures the complex dynamics of the environment. Subsequently, LLMs are employed to redefine the reward function and relabel the dataset, thereby establishing a well-defined Markov decision process based on the previously learned graph. This relabeling process ensures that the agent's decision space aligns with user intentions. Consequently, a predictive world model is obtained, offering insights into potential future states and facilitating graph-based planning. Moreover, by inputting the planned path of the agent in the graph-structured world model into LLMs, natural language explanations can be generated to provide transparency in the decision-making process. Experiments on the D4RL benchmark validated the effectiveness of our approach in long-horizon planning, its adaptability to different user tasks, and its inherent explainability.

Murotani, Kazuya; Hasejima, Noriyasu

doi: 10.1080/01691864.2024.2421920pmid: N/A

In recent years, studies on autonomous mobile robots for outdoor use have been actively conducted. In these studies, skid-steer robots such as crawler robots are often used. These mechanisms are not good at diagonally entering steps such as curbs, so it is necessary to consider the step entry direction for navigation of these robots. However, there are few previous studies that deal with detection of low height steps on the ground, path planning that considers the step entry direction, and tracking control, in a unified manner. In this study, we present a ground surface classification using LiDAR scan characteristics, a path optimization using quadratic programming with the step entry direction constraint, and a tracking control using TOPP to accurately follow the path. As a result of the experiment, we confirmed that it was able to detect 0.1 m height steps and climb over them by entering perpendicularly to the steps.

Aizawa, Koki; Kuramitsu, Yunosuke; Matsunaga, Hideki; Yasui, Yuji

doi: 10.1080/01691864.2024.2411685pmid: N/A

Efficient path planning is crucial for the safe autonomous operation of micro-mobility vehicles in unknown environments. When planning paths for micro-mobility, factors, such as the changes in the user's intended destination and user comfort, must be considered. Moreover, for driving in unknown areas, the path planner should not rely on predetermined detailed information like High-Definition maps (HD maps). In this paper, we propose an innovative path planning algorithm for automated driving that solely uses basic perception data from RGB camera, IMU and GPS, thereby eliminating the dependency on HD maps. We initially convert perception data into an occupancy grid map. Subsequently, a global path planner, based on a modified A* algorithm, computes an efficient trajectory, particularly adept to navigation in unknown environments and to scenarios where goal position moves. Additionally, we developed a local planner that optimize multi-clothoid curves using designed constraints to predict the best trajectory. Our experiments, conducted in both simulated and real-world environments, validate the effectiveness of our approach for micro-mobility path planning using only perception data.

Domae, Yukiyasu; Makihara, Koshi; Hanai, Ryo

doi: 10.1080/01691864.2024.2416015pmid: N/A

This study focuses on the opening width of parallel grippers in bin picking and proposes a method to optimize this width based on the evaluation of graspability in distance images. The method relies on fast graspability evaluation (FGE), which reduces the robot's grasp position calculation to a convolution problem between binary images representing the object and hand. This approach determines the hand parameters that yield the highest graspability score for the entire bin-picking scene. When introduced in multi-bin-picking tasks involving eight types of industrial parts, this optimization method demonstrated improvements in grasp success rates of up to 35% and an average improvement of approximately 14.4%. The proposed method confirms the importance of optimizing hand parameters in challenging industrial bin-picking scenarios.