In this paper, we propose a novel affordance model, which combines object, action, and effect information in the latent space of a predictive neural network architecture that is built on Conditional Neural Processes. Our model allows us to make predictions of intermediate effects expected to be obtained during action executions and make multi-step plans that include partial actions. We first compared the prediction capability of our model using an existing interaction data set and showed that it outperforms a recurrent neural network-based model in predicting the effects of lever-up actions. Next, we showed that our model can generate accurate effect predictions for other actions, such as push and grasp actions. Our system was shown to generate successful multi-step plans to bring objects to desired positions using the traditional A* search algorithm. Furthermore, we realized a continuous planning method and showed that the proposed system generated more accurate and effective plans with sequences of partial action executions compared to plans that only consider full action executions using both planning algorithms.

This article presents a method for guaranteeing the safety of a system with both parametric and nonparametric uncertainties, while at the same time decreasing the conservatism compared to existing approaches. This is obtained by combining robust adaptive control barrier functions (RaCBF) and Gaussian process control barrier functions (GPCBF). We provide a condition under which the considered system is safe with a given probability, and show that the proposed method is less conservative than GPCBF. We evaluate the method through a simulation study, where we consider a force controlled robot manipulator in contact with a partially unknown environment. The results show that our proposed GPRaCBF can guarantee bounds on the contact forces despite parametric and nonparametric uncertainties in the contact dynamics and outperforms GPCBF in terms of the conservatism.

Bilateral movement has been reported to enhance actual motor performance more than unilateral movement and increase motor cortex activation more than unilateral movement. We hypothesized that bilateral motor imagery would result in greater improvements in event-related desynchronization (ERD) amplitude and ball rotation performance than unilateral motor imagery. Forty healthy young adults participated in this study. The mental practice session was performed in the unilateral hand condition (right and left hands) and bilateral hand condition for four blocks. ERD in mental practice was calculated using electroencephalography (EEG). The ball rotations were compared between the pre-and post-tests of mental practice. The results showed a greater improvement in actual motor performance with unilateral motor imagery than with bilateral motor imagery. ERD analysis showed no significant differences in multiple comparisons. These results did not support this hypothesis. Nonetheless, the present study showed that unilateral movements resulted in greater performance gains than bilateral movements, suggesting that bilateral motor imagery may have additional control processes similar to actual movements.

In this study, we present a method for moving a snake robot in a pipe while generating slips between the robot and the pipe wall. This method is intended to be applied to pipe cleaning. The structure of the robot consists of a helix and reverse helix connected by a passive roll joint, and the rolling velocities of the two helixes are controlled differently to generate slips. We designed the target paths to ensure that the robot could clean the entire pipe wall even if the brushes were attached while spaced out. A feedback control that adjusted the rolling velocities of the two helixes to follow the target path was introduced. The position of the snake robot can be controlled to move independently in and around the pipe axis. We verify the effectiveness of the method using simulations and experiments. We confirmed that by using feedback control, the robot could roll according to the target path, but only in the backward direction of the robot. We also confirmed that applying this method to cleaning the entire pipe wall is difficult because of the different errors of each brush.

ABSTRACT This work presents the first results using Smoothed Particle Hydrodynamics (SPH), a mesh-free technique, to simulate underwater vehicle motion with the goal of achieving sufficient physical realism and computation time performance capabilities. The objective is not to get very accurate values for the hydrodynamic parameters, but to show that SPH can simulate hydrodynamic parameters with the same order of magnitude as the reference, in order to allow a realistic control of robots in water. First, spherical objects are simulated to check buoyancy realism, speed limit existence, and hydrodynamic parameters in comparison with reference values. Then, horizontal and vertical movements of a capsule-shape object and a real torpedo-shape underwater robot are compared. The results show that buoyancy is respected, and that spherical objects reach a speed limit in accordance with the laws of physics. In addition, added-mass is simulated with 20 % variation on average with respect to the reference and varies homothetically with respect to the object's size. In contrast, drag forces cannot not be simulated with the same level of realism without reducing the particle size, which makes the simulation last longer. SPH for underwater robotics simulation appears to be promising, and ways of further improvements are being considered.

Dialogue systems implemented with android robots are expected to provide not only advanced conversational ability but also reliability and hospitality due to their human-like appearance. In this study, we aim to develop a hospitable dialogue system by encouraging open-ended utterances and responding adaptively to give users the feeling of being heard. To achieve this, utilizing large language models (LLMs) is a promising option, but task-oriented dialogue systems implemented with only LLMs often generate irrelevant, inconsistent, or contradictory utterances. Therefore, we propose a scenario-based dialogue system that subdivides the task into smaller sub-tasks, such as summarization, information extraction, and response generation, and uses LLMs in a fine-grained manner to overcome such shortcomings of the LLM-based dialogue system. Our system was evaluated in the tourist-spot recommendation task of the Dialogue Robot Competition 2022 and achieved second place in the preliminary round and first place in the final round, outperforming other rule-based dialogue systems. However, we also identified several challenges when using LLMs for android dialogue systems, including response delays due to computational complexity, hallucinations, and coordination issues between generated utterances and robot control.

In this paper, we present a multimodal dialogue system combining a neural response generation mechanism, a reranking mechanism, and a rule-based avatar control mechanism. Our system was submitted to the open track at the Fifth Dialogue System Live Competition and won second place. Remarkably, our system received the best human evaluation performance for visual information control (i.e. speaking style of avatar) in the preliminary round. The assessment of the competition evaluators revealed that our system generates natural utterances appropriate to the conversational context and topic with an appealing speaking style. Through the analysis, we found that our devices, such as post-processing on speech recognition results and the final response selection method, are effective, but we also found room for improvement, such as speech recognition errors and challenges in the reranking module.

This paper proposes a novel optical proximity sensor skin that can cover entire robot links and has high responsiveness, wire-saving properties, and suitability for covering narrow surfaces. Moreover, it can extract information that can help to control robots. Although implementing these characteristics simultaneously has been problematic for sensor skins covering links, the proposed sensor skin achieves this using a newly developed analog computing circuit. This analog computing circuit only consists of several resistors and diodes and is based on the splitting ratio of photocurrents variation based on the design of the resistance values. If the output from the proximity sensing element is interpreted as a virtual force, the processed output denotes its magnitude and the location of the point of action. Even if multiple objects are detected, the outputs are associated with the resultant virtual force because photocurrents are merged. Therefore, the characteristics such as high responsiveness are maintained. This work fabricated a prototype to evaluate the proposed sensor skin. Then, collision avoidance experiments against human hands were conducted to validate the efficiency of the proposed sensor skin for generating robot motions. These results verified that the proposed sensor skin is suitable for robots reacting quickly to unknown environments.

Quadrupedal animal, such as dogs, typically have soft but heavy abdomens, whose swinging motion significantly contribute to the energy efficiency of their entire locomotion. This study examines the effect of a swinging mass on quadrupedal locomotion by simulating the soft abdomen of its biological counterparts. A quadruped robot is designed with a swinging mass beneath its trunk. Its pitching motion was evaluated by a mechanical model. The experiment results demonstrate that the swinging mass effectively reduces the pitching motion, thereby improving the energy efficiency of the robot. Although the swinging mass does not always guarantee the energy efficiency of locomotion, the spring stiffness that supports the mass can be obtained for energy-efficient locomotion.

Extraction of low-dimensional latent space from high-dimensional observation data is essential to construct a real-time robot controller with a world model on the extracted latent space. However, there is no established method for tuning the dimension size of the latent space automatically, suffering from finding the necessary and sufficient dimension size, i.e. the minimal realization of the world model. In this study, we analyze and improve Tsallis-based variational autoencoder (q-VAE), and reveal that, under an appropriate configuration, it always facilitates making the latent space sparse. Even if the dimension size of the pre-specified latent space is redundant compared to the minimal realization, this sparsification collapses unnecessary dimensions, allowing for easy removal of them. We experimentally verified the benefits of the sparsification by the proposed method that it can easily find the necessary and sufficient six dimensions for a reaching task with a mobile manipulator that requires a six-dimensional state space. Moreover, by planning with such a minimal-realization world model learned in the extracted dimensions, the proposed method was able to exert a more optimal action sequence in real-time, reducing the reaching accomplishment time by around 20%.