Fish Swarm: Hydrodynamic Interactions of Robotic Fishes

Fish Swarm
Author’s Name
Institutional Affiliation
City and Country’s Location
Email Address
Date
Fish Swarm
One of the areas that have attracted scientists’ interests in the field of engineering, physics, and biology is how robotic fish move in groups. The potential advantage of swimming together in well-coordinated schools is to enable fish to extract energy from their neighbors’ vortices to reduce their locomotion costs [1]. When engineers are making robotic fish, they emphasize how swimming together can save energy. Fish swarm has algorithms that allow them to swim in groups, and experiments have revealed that group swimming makes robotic fish stable underwater. However, one of the primary challenges that engineers encounter is the lack of an effective bottom-level control system for robotic fish. The paper focuses on how fish swarm work, the technique used to control fish auto swim in a group, and Fumin Zhang’s fish swarm and the technique used to control them. Additionally, it determines whether any group succeeded or failed in using ultrasonic or sonar to locate other fish, the techniques used underwater for locating, and their advantages and disadvantages.
How Fish Swarm Work
Engineers are working hard to promote the robustness and diversity of locomotive strategies of robotic fish in fluid environments. Biologists believe that fish swarm uses the Central Pattern Generator (CPG) to generate rhythmic outputs that control their locomotion [2]. In that light, robotic fish have CPG controllers with feedback, reinforcement learning, and proprioceptive sensing. The CPG controller helps the robotic fish to adjust their bodies’ undulation when they receive feedback from proprioceptive sensors, which are decoded through reinforcement learning. In addition, schooling fish dynamically optimize body undulations so that they can save energy and improve their swimming efficiency without taking into account their spatial formations [3]. Lateral line and vision are the two primary sensory modalities that fish swarm use to gather environmental information so that they can make effective movement decisions. However, researchers identified that fish swarms can still react to nearby flows even when their vision systems and lateral line are impaired. Fish use proprioception sensors to know the local flow dynamics and adjust to kinematics appropriately. Besides, proprioception means muscle force and self-motion sensing.
Technique Used to Control Fish Auto Swim in Group
Neural networks and hydrodynamic interactions are used to control fish auto swim in groups. Robotic fish swim side-by-side, which makes them generate rhythmic signals that facilitate swimming efficiency in groups [1]. A robotic fish has a camera, motor, and battery. The motor propels the fish, creating a side-by-side motion. An acoustic communication system resembles the CPG controller, and it helps fish change their swimming speed and control their buoyancy. The diagram below shows how hydrodynamic interactions control fish auto swim in a group [3].
Figure 1
How Hydrodynamic Interactions Control Fish Auto Swim in a Group


Source: https://iopscience.iop.org/article/10.1088/1748-3190/ac165e/pdf
Figure 1(a) shows how robotic fish generate hydrodynamic interactions by flipping their tail tip side-by-side. Part (b) shows the graph of the movement generated by the fish group swimming. In particular, the hydrodynamic interactions enable the robotic fish to respond to the flow fields underwater, generating movement. Fish swarm can optimize their body undulations, making it easy to control themselves. Since robotic fish have sensors, they respond to each other movements, meaning that they can increase or reduce their swimming speed automatically.
How Fumin Zhang’s, from Georgia Institute of Technology, Fish Swarm Works
Zhang specializes in electrical engineering, and he is among renowned engineers who have studied fish swarm locomotive strategies. His research deals with mobile sensory networks. The engineer significantly contributes to the designing of sensing, communication algorithms, and control for spatial-temporal stochastic and mobile sensing agents. In particular, he is involved in data collection and marine robotics. Zhang has significantly contributed to the study of autonomous underwater vehicles (AUVs). The scientist uses a fleet of aquatic robots in performing distributed sampling underwater. Zhang’s fish swarm uses acoustic networks to enhance effective functionality. Robotic fish comprises an algorithm that facilitates underwater signal communication to coordinate the fish swarm [4]. In that light, the fish swarm from the Georgia Institute of Technology can achieve specific tasks together by using the open-source, community-shared, and open-architecture infrastructure. However, the significant challenge of using acoustic communication underwater is the limited bandwidth and range. Underwater communications of Zhang’s fish swarm are enhanced by ROS and MOOS-IvP. As a result, Zhang is focused on improving underwater robotics so that they can work more effectively.
Zhang’s Technique to Control Fish Auto Swim in Group
Zhang’s robotic fish uses algorithms to simulate swimming in groups. These underwater robots have sensors and motors that are powered by a battery to work effectively. Although acoustic communications have limited bandwidth and range, it works for fish swarm since they swim in close proximity [5]. The fish swarm can reduce or increase their swimming speed uniformly. For instance, when one fish decreases its swimming speed, others around it sense that and respond accordingly. The robotic fish senses movement in any direction, including up and down. Since Zhang and colleagues have in-depth knowledge of underwater robotics, they emphasize acoustic communication to facilitate group swimming without collision. The sensors enable the fish to adjust their speed based on the flow of fluids. The effectiveness of fish swarm group swimming is crucial for Zhang and colleagues since it helps them gather detailed information about the marine environment. The capability of these robotic fish to swim up and down in large water bodies, such as seas and oceans, is essential for data collection. Consequently, Zhang emphasizes the robustness of group swimming through acoustic communication that enables fish swarm to operate automatically.
The Success or Failure in Using Sonar or Ultrasonic to Locate Robotic Fish
Several groups succeeded in using ultrasonic as a strategy for locating robotic fish in a swarm. Nevertheless, the only disadvantage of the ultrasonic method is its limited bandwidth coverage underwater. Specifically, Zhang and colleagues managed to use ultrasonic to locate fish swimming in a group. Another group that succeeded in using this method is Li et al. (2020). These researchers believe that schooling fish must coordinate their movements to swim together. In that light, the tail beat is facilitated by the motor and creates waves in the water. Robotic fish have sensors that enable them to communicate through acoustic communication and respond accordingly. The propulsive motion of fish swimming in groups creates a regular alternating vortices pattern that enables robotic fish to move together in a specific direction [6]. Mi et al. (2017) use a teleoperation framework that reveals how gesture recognition enhances communication among animals [7]. Notably, the group succeeded to show that underwater robotic fish use gestures created through the ultrasonic method to interact so that they can achieve similar objectives. Without effective communication among robotic fish, they cannot swim together or coordinate effectively.
Tian et al. (2020) argue that robotic fish are poor group swimmers. As such, this group failed to use ultrasonic to facilitate group swimming of robotic fish. The primary reason these researchers failed is due to the difficulty of tuning the motion parameters that enable fish to communicate with each other [8]. The group tried to use a computational fluid dynamics (CFD) simulation and did not manage to regulate the motion parameters of the robotic fish swarm. The complex fluid environment requires a high battery capacity, which was a challenging goal to achieve for the group. In particular, a fish swarm requires a high level of coordination in group swimming. Using unreliable batteries might adversely affect the motor and tail beat frequency, causing collisions of fish in the swarm [9]. An irregular movement or swimming speed of a robotic fish in a swarm influences others. For instance, a collision among robotic fish might occur or a halt of group swimming altogether.
Li et al. (2021) used bio-inspired robotic fish to determine how fish swarm group swimming can save energy to facilitate ultrasonic communication [10]. Researchers found a correlation between group swimming and saving energy. Three-dimensional computational fluid simulations reveal that vortices from the tail tip of robotic fish control the speed and direction of fish. During group swimming, the fish swarm generates reson...