Artificial Life and Robotics最新文献

In this study, we propose a centipede-like swarm robotic system capable of utilizing “unfavorable” environmental effects such as robot–environment collisions, robot–robot collisions, signal noise, and signal interval, to aid navigation and exploration in 2D unknown environments. Traditional swarm robots demand complex systems to counteract “unfavorable” environmental effects, making the design of robots to handle unknown environments with theoretical infinite possibilities exceedingly challenging. In contrast, our approach does not view these “unfavorable” environmental effects as foes to be combated, but instead employs strategies like implicit control to convert them into beneficial influences on the task. This simplifies the swarm robot system, enabling navigation and exploration with only goal direction as the single input, eliminating the need for inter-robot observation, inter-robot communication and obstacle detection. Finally, we validate our system’s effectiveness through simulations in the CoppeliaSim environment, demonstrating the benefits of “unfavorable” environmental effects and the method’s universal applicability.

Currently, the phenomenon of ocean desertification, known as rocky shore denudation, is occurring all over the world. One of the causes is the feeding damage by sea urchins on seaweed beds called seagrass beds. To restore the seaweed beds, it is necessary to exterminate the proliferating sea urchins. We have developed an automated survey system that provides information on sea urchin habitat distribution and seafloor topography to streamline the extermination process. Based on a small surfboard-based ASV developed in our laboratory, an underwater camera and sonar are attached to the ASV (autonomous surface vehicle), which navigates automatically to collect data. By analyzing the obtained data, we were able to obtain the habitat distribution of sea urchins and a topographic map of the seafloor, which are necessary for sea urchin extermination.

In the swarm robotics community, Lévy walk has been recognized as one of the most efficient search strategies for the environment, with sparse targets that robots have no prior knowledge of. Generally, Lévy walk is generated by following the Lévy distribution. Our previous results also confirmed that the Lévy walk outperformed the usual random walk for exploration strategy in real swarm robot experiments. On the other hand, it has been reported in several papers that each individual in swarm robots does not follow Lévy distribution due to collision avoidance from other robots, resulting in inefficient search. Therefore, we introduced concessions to the swarm robots to improve search efficiency. This paper investigated the performance of the Lévy walk with concession. Robots concede other robots when they receive the signal that other robots execute longer walks. We conducted a series of computer simulations varying ranges detecting other robots’ walk distance signals, the number of robots, the number of targets, and the distribution of targets. The results suggest that the search efficiency of Lévy walk was improved by concession. Furthermore, we confirmed that improving search efficiency saturates beyond the threshold of range detecting other robots’ walk distance signals.

Balloon robots consist of a balloon body, which is filled with helium to provide buoyancy. In particular, fish-type balloon robots (FBRs) incorporate caudal- and pectoral-fin motions as their propulsion mechanism, which can be combined to realize complex motions. However, the propulsive force generated by the pectoral-fin motion is less, which is a disadvantage, and it is necessary to increase the propulsive force. Since the weight of FBR is limited, it is crucial to select a fin that can generate larger propulsive force at the same weight. In this study, pectoral fins with different shapes and materials are developed, and the propulsive force generated by various fin movements is measured. Finally, the results are compared and discussed relative to the influence of shape and softness on the propulsion.

Maritime Autonomous Surface Ships (MASS) are expected to be used in logistics, transportation, security, etc. in the future. When operating a ship, we have to pay attention to invisible objects under the water. Objects above the water surface can be detected by the camera system. On the other hand, underwater objects are generally detected by acoustic sensors called “sonar”. Almost all sonar products are expensive, so it is not realistic to mount them on all MASS. Therefore, surveying underwater with a survey ship and preparing charts that includes underwater objects, it realizes safe and reasonable navigation of MASS. Since mobility systems for tourism and transportation are required in urban areas, MASS is one of the solutions for those systems in urban rivers. However, it is difficult surveying in urban rivers. This paper describes some experimental results, technical issues, and the solution for underwater surveying of urban rivers, for which the demand for MASS is expected to increase in the future.

Basic analog–digital motion detection circuits with low power consumption were proposed based on the vertebrate retina. The motion sensor based on the retina is constructed with a one- or two-dimensional array of the unit circuits. The proposed unit circuit for motion detection is constructed with an analog circuit for photoelectric conversion and the digital circuit for generating the motion signal. The metal oxide semiconductor (MOS) transistors utilized to the analog circuits are operated in the subthreshold region. The analog circuit has the characteristic of the low power consumption. The proposed circuit was evaluated by the simulation program with integrated circuit emphasis (SPICE) with the 0.6 μm complementary metal oxide semiconductor (CMOS) process. The test circuits of basic digital circuits were fabricated with the same process. In the simulation and the experiment, the power supply voltage was set to the low voltage. We found that the digital circuit becomes low power consumption because the MOS transistors was operated in the subthreshold region by setting the low voltage. The proposed circuit is characterized by the simple structure and the low power consumption. In the future, the novel motion detection sensor with low power consumption can be realized by applying the integrated circuits constructed with the array of the proposed unit circuits.

This paper focuses on generating collective behavior of a robotic swarm using an attention agent. The selective attention mechanism enables an agent to cope with environmental variations which are irrelevant to the task. This paper applies attention mechanisms to a robotic swarm for enhancing system-level properties, such as flexibility or scalability. To train an attention agent, evolutionary computations become a promising method, because a controller structure is not restricted by a gradient-based method. Therefore, this paper employs a deep neuroevolution approach to generating collective behavior in a robotic swarm. The experiments are conducted by computer simulations that consist of the Unity 3D game engine. The performance of the attention agent is compared with the convolutional neural network approach. The experimental results showed that the attention agent obtained generalization abilities in a robotic swarm similar to single-agent problems.

In recent years, there has been a growing demand for drivers’ drowsiness detection technology in driver monitoring systems for level 3 automated vehicles. In a previous study, sparse modeling was applied to the facial skin temperature distribution which can be remotely measured with a thermography camera, achieving an estimation accuracy of about 74% in three stages. However, the constructed drowsiness estimation model had low generalization performance because it was an individual model, and the sparse modeling had the property of easily decomposing behavioral indicators other than drowsiness. Therefore, in the construction of a general model using sparse modeling, individual features among subjects may be preferentially decomposed, and a method to absorb these effects is needed. Thus, in this study, we devised an attempt to reduce the influence of decomposition of individual features by applying averaged faces. Partial averaging of images across subjects is thought to remove such effects. In this study, we attempted to construct a general model for drowsiness estimation by applying sparse modeling to facial thermal images averaged across subjects, and to examine the possibility of using averaged facial thermal images in constructing general model. As a result, we obtained an estimation accuracy of approximately 54.6% in average by applying averaging, which is 7% higher than that using the original images, and succeeded in reducing its standard deviation by 4–6.9%. As the result, modeling with averaged images was shown to be effective in improving generality.

Safety–critical system is important in a human–robot collaborative environment. Control Barrier Functions (CBFs)-based methods have emerged as a practical tool for the safety–critical control of autonomous systems. The design of CBFs is difficult to tune. Also, once additional constraints are introduced, the quadratic programming may encounter infeasibility. This paper proposes a safety–critical controller based on a control barrier function using a quasi-saturation function. To avoid infeasibility, we propose to separate the tracking controller from the safety controller. To facilitate the design of the control barrier function, we also propose the control barrier function using the quasi-saturation function. Numerical simulations are presented to show the effectiveness of the proposed method.

This paper deals with the design of a motion and force control scheme for underwater vehicles, each of which has a manipulator. For a subsea operation that requires a contact between the manipulator’s tip (e.g., the hand) and various types of environments, it is desirable that the mechanical impedance of the manipulator is adjusted according to the contact surface. From this point of view, the paper focuses on the design of an impedance control scheme. Several impedance controllers have been developed. Most of them were designed on the assumption that the control capability of the vehicle is the same as that of the manipulator. However, it has been pointed out in the literature that for a real underwater robot, its vehicle control is more challenging than its manipulator control, because the vehicle has much larger inertia, and many more inaccurate position sensors and actuators than the manipulator. In view of this fact, we develop an impedance control scheme for the manipulator under the condition that the vehicle is independently controlled by a motion controller with poor performance. Moreover, we provide the results of simulations for comparing an existing controller with the proposed one.