Verification of Acoustic-Wave-Oriented Simple State Estimation and Application to Swarm Navigation

Cooperative swarming behavior of multiple robots is advantageous for various disaster response activities, such as search and rescue. This study proposes an idea of communication of information between swarm robots, especially for estimating the orientation and direction of each robot, to realize decentralized group behavior. Unlike the conventional camera-based systems, we developed robots equipped with a speaker array system and a microphone system to utilize the time difference of arrival (TDoA). Sound waves outputted by each robot was used to estimate the relative direction and orientation. In addition, we attempt to utilize two characteristics of sound waves in our experiments, namely, diffraction and superposition. This paper also investigates the accuracy of state estimation in cases where the robots output sounds simultaneously and are not visible to each other. Finally, we applied our method to achieve behavioral control of a swarm of five robots, and demonstrated that the leader robot and follower robots exhibit good alignment behavior. Our methodology is useful in scenarios where steps or obstacles are present, in which cases camera-based systems are rendered unusable because they require each robot to be visible to each other in order to collect or share information. Furthermore, camera-based systems require expensive devices and necessitate high-speed image processing. Moreover, our method is applicable for behavioral control of swarm robots in water.

A GPU-accelerated image reduction pipeline

We developed a high-speed image reduction pipeline using Graphics Processing Units (GPUs) as hardware accelerators. Astronomers desire to detect the emission measure counterpart of gravitational-wave sources as soon as possible and to share in the systematic follow-up observation. Therefore, high-speed image processing is important. We developed a new image-reduction pipeline for our robotic telescope system, which uses a GPU via the Python package CuPy for high-speed image processing. As a result, the new pipeline has increased in processing speed by more than 40 times compared with the current one, while maintaining the same functions.