Robots are becoming essential in human environments, requiring them to behave in a socially compliant manner. Although previous learning-based methods have shown potential in social navigation, most have treated pedestrians as individuals, failing to account for group level interactions. Additionally, these methods have modeled pairwise interactions only in the spatial domain, overlooking the temporal evolution of relations among agents. In this letter, the above limitations are addressed by proposing a novel spatio-temporal graph attention network that explicitly models group level interactions in both spatial and temporal domains. Specifically, a novel group-awareness mechanism is designed to model group-aware behaviors, and a new network is proposed to capture spatio-temporal features of relations among agents while leveraging the model-free deep reinforcement learning to optimize the group-aware navigation policy. The test results show that our approach outperforms the baselines in all metrics in both simulation and real-world experiments. Furthermore, quantitative analysis of questionnaire responses further verifies the benefits of our method in group awareness and social compliance.

We propose two motions for a snake robot to move through crowded pipes. One motion is an S-shaped twisting motion performed by the snake robot as it moves around crowded pipes in parallel to the pipes. To develop the shape of the snake robot, the S-shaped twisting motion creates large and small semicircles alternately. In the S-shaped twisting motion, lateral rolling is used to generate the robot's twisting motion along its body to maneuver among crowded pipes. The other motion is an S-shaped traveling wave motion used by the snake robot to move among crowded pipes perpendicularly to the pipes. The S-shaped traveling wave motion adds a hyperbolic shape to the snake robot by utilizing a hyperbolic function. The shape of the hyperbolic function produces a little wave in the S-shaped traveling wave motion, and the wave is transferred from the tail to the head to move ahead. In this study, we describe the design of these two motions and exhibit the results of trials in simulation and with a real snake robot to validate the effectiveness of our proposed snake robot's motion. GRAPHICAL ABSTRACT

The movement capabilities of snake robots allow them to be applied in a variety of applications. We realized a snake robot climbing in crowded pipes. In this paper, we implement a sinusoidal curve control method that allows the snake robot to move faster. The control method is composed of a hybrid force–position controller that allows the snake robot to move more stably. We conducted experiments to confirm the effectiveness of the proposed method. The experimental results show that the proposed method is stable and effective compared to the previous control method that we had implemented in the snake robot.