Artificial Life and Robotics最新文献

This study explores the influence of material selection scenarios on the evolutionary design of soft robots. Soft robots have vast potential applications that rigid robots cannot accomplish due to their lightweight and flexible nature. Despite rapid development of soft robotics in recent years, its effective automatic design method remains undeveloped. To combat this, evolutionary computation has been proposed for automatic soft robot design. Material selection has been identified as crucial in influencing the outcomes of autonomously designed soft robots, yet understanding of its impacts remains limited, particularly in changing material selection scenarios across generations. In this study, we hypothesized that these changing scenarios could influence the evolved robots. Initial experiments, based on travel distance and energy efficiency, suggested material selection could greatly impact the final robot. Further experiments, adjusting material ratio and rewarding specific material use, identified how certain material selection scenarios led to changes in the structure and performance of the evolved soft robots. These findings contribute to automatic design methodologies in soft robotics.

In randomized clinical trials, we assumed the situation that the new treatment is not adequate compared to the control treatment as a result. However, it is unknown if the new treatment is ineffective for all patients or if it is effective for only a subgroup of patients with specific characteristics. If such a subgroup exists and can be detected, the patients can receive effective therapy. To detect subgroups, we need to estimate treatment effects. To achieve this, various treatment effect estimation methods have been proposed based on the sparse regression method. However, these methods are affected by noise. Therefore, we propose new treatment effect estimation approaches based on the modified covariate method, one using lasso regression and the other ridge regression, using the (L_0) norm. The proposed approach was evaluated through numerical simulation and real data examples. As a result, the results of the proposed method were almost the same as those of existing methods in numerical simulations, but were effective in real data example.

In today’s information-rich society, the importance of recommender systems for matching items and customers is increasing day by day. The development of e-commerce sites and review sites has made it possible to access a large amount of product descriptions and user reviews, and it is believed that more advanced recommendation models can be proposed by efficiently utilizing this text information. ConvMF is the first model that integrates text and probabilistic matrix factorization(PMF) which is one of the matrix factorization methods. In this method, features are extracted from item text such as item descriptions using CNN architecture and integrated into PMF. However they focus only on the item text and not on the user factor. As a result, this method can not reflect user characteristics. Therefore, this paper proposes a new recommender system to extract both item and user features from item and user text using CNN and integrate them into matrix factorization.

In vehicles with active suspensions, if handling stability is improved strongly by using a LQ controller for a road surface such as large bumps, ride comfort may deteriorate even for a road surface such as not so large bumps. Recently, to avoid the problem, a nonlinear active suspension control scheme has been proposed. However, in the proposed nonlinear controller, it is required that vehicle mass does not vary. In practice, vehicle mass varies greatly. If vehicle mass varies, the controller has to be redesigned. In this paper, to address the problem, based on the proposed nonlinear control scheme, we develop a new robust active suspension control scheme.

This study proposes a vacuum gripper that can measure the effective suction force applied to an object using a 6-axis force sensor. The object falling in a vacuum gripper occurs when a gap between the object and the pad no longer allows the negative pressure in the chamber. This is rephrased as the absence of a compression force between the object and the pad. Therefore, the force distribution that the attached object pushes the pad during suction reflects the external force, i.e. the effective suction force, required to remove the object. To confirm the validity of the developed vacuum gripper, we conducted an experiment where the developed vacuum gripper sucks a flat plate. As a result, we confirmed that the 6-axis force sensor’s signals mean the effective suction force by applying an external force to remove the place.

Human augmentation technologies can strengthen and compensate for the lack of human abilities associated with aging and physical disabilities. Efforts have been made to apply these technologies in manufacturing and agricultural industries to improve work efficiency, reduce fatigue, and cover differences in workers’ skills. This study proposes a human augmentation hand to assist users in performing intelligent tasks requiring high brain functions, such as cognition, planning, judgment, and memory. Solving a dissection puzzle is used as an example of an intelligent task. The human augmentation hand has a view of the puzzle blocks with an attached camera, and the puzzle blocks required for solving the dissection puzzle are derived using a full-search algorithm. The system can then provide user hints for solving puzzles and assist in handling puzzle blocks. An experiment is conducted to confirm the operation of the system and examine its usefulness. A National Aeronautics and Space Administration task load index (NASA-TLX) questionnaire is used to investigate the subjective workload perceived by users. The experimental results reveal that the proposed system improves task efficiency and reduces workloads that require high brain functions.

The human postural control function is determined by inputs from three senses, namely, vision, balance, and somatosensation. Many studies have attempted to induce body movements in arbitrary directions using external stimuli to maintain and improve postural control functions. Previous studies demonstrated that sensory information from the plantar region of the foot is involved in the understanding of front–back position and regulation of body movements. In the present study, we investigated a method to induce postural control and walking movements in an arbitrary direction by presenting matrix-shaped tactile stimuli (MSTS) to the plantar region of the foot. We developed a tactile stimulation device to investigate body sway when the MSTS was presented to the dorsal surface region of a foot in a standing posture. This device is envisioned to be used as a wearable guidance tool in the future. In addition, the measurement experiment used a high-speed camera and an accelerometer placed on the top of the subject’s head to track the subject’s movements in response to the MSTS presented by the device. In this paper, the relationship between the control parameters (L, T) of tactile stimulation MSTS and body sway and its direction was considered. The results showed that the introduction of MSTS on the sole of the foot could induce body sway. The results also suggest that it is possible to induce body sway in the respective expected directions by setting appropriate parameters (L, T).

In this paper, we propose an algorithm for estimating respiratory state using near-infrared facial video images. Estimation of respiratory state is an important indicator for early detection of respiratory diseases. In particular, there is a demand for monitoring respiratory state during the night. One method of monitoring respiratory state is to use contact-type sensors. However, this method requires the installation of many sensors and a visit to a hospital, which place a burden on patients. Therefore, we propose to acquire respiratory-induced features from near-infrared face video images and investigate their similarity to measurements obtained with a respirometer for non-contact monitoring of respiratory state in the dark. Respiratory-induced features were obtained from pulse wave signals extracted from the face video images. The results showed correlations in several respiratory states. This study opens some perspectives in non-contact monitoring of respiratory states.

Dynamic earphones/headphones and microphones can detect sounds below the human audible frequency. Based on this principle, we propose a new method of human pulse measurement using consumer earphones and headphones. It is a simple signal separation-based method utilizing pressure changes inside the ear canal and around the tragus caused by heartbeats. As a result of feasibility evaluation using an electrocardiogram, the pulses derived by the earphone/headphone were highly feasible in terms of accurate peak-to-peak determination. Furthermore, we estimated the frequency characteristics of the audio devices at the frequency of interest, the center frequency of the heart rate (around 1.4 Hz), which enables us to reproduce the original, non-distorted pulse waveform. Although this is an entirely different methodology from photoplethysmography, it is promising because heart rate can be measured while listening to music.

The authors have been studying robots equipped with pulse-type hardware neuron models (P-HNMs) that mimic biological neurons. In a previous study, we developed a quadruped robot system mounting P-HNMs. The authors confirmed the generation of animal-like gait that transitions gait in response to the speed of movement. However, significant circuit variations occurred between the circuit boards of the P-HNMs because the discrete elements had been soldered to the circuit board by hand. Therefore, the circuit characteristics have to adjust using variable resistance for each experiment. In this paper, the authors developed an integrated circuit of P-HNMs to reduce the circuit variation. Also, we mounted the integrated circuits of P-HNMs on a quadruped robot system. As a result, we confirmed that an integrated circuit of P-HNMs does not require any adjustments to actuate the robot system. Also, the author established that the proposed quadruped robot system generates a gait like the “Walk” and the “Trot” of quadruped animals through the walking experiments. The generated gaits are three to seven times longer than the previous robot system.