Will a 'Plastivorous' Fish be Created to Help Clean Ocean?

We are a student group studying at the university of bristol and majoring in bio-robotics. Since ocean micro-plastic pollution become a world-class problem, we plan to design and fabricate a 'plastivorous' robotic fish to help. We look forward to your voice and suggestions for common progress, so please feel free to contact us through our blog website.

World-class problem & What can we do

Plastics are endurable and over-produced. That means plastics are hard to be decomposed to natural elements, but simply divided into smaller and smaller parts, and finally, be a part of the environment as toxic monsters. Those small parts are so-called microplastics (diameter < 5 mm), that are almost everywhere. They have entered the food chain, shown up in animals' nests and become destitute children's toys, so that aroused attention from all walks of life.

Current solutions for this world-class problem can be roughly classified as sampling, detecting, recycling, degrading and so on by different processing stages. In this project, we will concentrate on the sampling process, since the remainder of the parts involves lots of chemistry or biology techniques which are high beyond our abilities.

Existing solutions for micro-plastic sampling

Sampling is the basic process for further research. There are generally two approaches to sampling the microplastic concentration in seawater: net-based sampling methods and bottle intake sampling methods. [5] The former deploy nets of different mesh sizes at the 1 back of a ship or vessel, drag the nets forward for some time at a certain water depth, and obtain microplastic samples under the corresponding water volume after filtering through the nets. Commonly used nets include bongo nets (>500 µm), manta nets (>300 µm) and plankton nets (>200 µm and >400 µm), and the capacity to filter microplastics is restricted by the size of the mesh. [5]

    However, according to a series of comparative studies in [5], the level of microplastic concentrations calculated by net-based sampling approaches were three orders of magnitude lower than those reported using bottle intake sampling methods. In addition, the number of fibres in samples obtained with bottles was considerably higher than that with nets. Since the volume of sampled seawater collected with a bottle can be accurately measured, the sampling error introduced in the bottle intake sampling methods is small, which can help improve the reliability of the data set. Considering no requirement for expensive specialized equipment and the repeatability of experiments, the bottle sampling method is a promising technique to realize the collection of marine microplastics.

What is 'Plastivorous'?

We use 'plastivorous' to imply our robotic fish intakes sea water with microplastics for sampling, just as an increasing number of marine creatures do now (So please use fewer plastics products in the future for our common earth's sake). This strategy is pretty close to the bottle intake sampling method discussed above but uses a robot as a carrier to sample seawater autonomously. If control the cost and achieve stable performance, the bio-robotic solution may save researchers time and money. Moreover, considering robotic fish as a platform for scientific research, it has functional extensibility. More discussions will be unfolded later in 'Limitations and Future Outlook'.

Could it be possible?  (Similar existing robots)

Using robotic fish to sample seawater requires the 'plastivorous' fish can swim in the ocean along a 3D trajectory and avoid obstacles autonomously without turning its belly. In this section, we take a closeup of each of the techniques and probe the project's feasibility.

Propulsion and pitch motion

In nature, approximately 85 of fish use the Body and/or Caudal Fin (BCF) as the main
propulsion mode to swim forward. BCF swimmers continuously flap their bodies to generate a travelling wave increasing amplitude from head to [6].  We analyzed some underwater robots and extracted the main mechanisms listed below.

Forward and turn

Depth control

Self-balancing

    Since the motion mode of the robot is controlled by the bionic fishtail, it is urgent to solve the self-stabilization ability. Kanjanawanishkul proposed a self-balancing bicycle using LQR and PID methods 8, which completed a similar way of self-balancing underwater.
    In the structure of this self-balancing system, the rotating acceleration of the momentum wheel will give the entire peripheral system a self-rotating torque, thus realizing the self-balancing function. The inputs of the control system are the rotation angle and speed of the momentum wheel, the error deflection angle of the balanced attitude between the peripheral system and the target, and the change rate of the deflection angle.

    Inspired by the self-balancing bicycle, the self-balancing system of the underwater robot will also be equipped with a momentum wheel to achieve self-balancing. This structure requires the robot to have a sensor for balancing the instrument and a servo motor that can record the acceleration of the momentum wheel. According to analysis from Nasir 9, LQR has lower energy consumption and is more stable. In the same steady state, PID may have more severe jitter than LQR.

Obstacle Avoidance

Computer Vision

A popular way to realize obstacle Avoidance is computer vision, which depends on fair ambient light and is relatively expensive and heavy.

Sonar System

    Braginsky proposed a method of obstacle avoidance employing a sonar system. This is an underwater obstacle avoidance system composed of horizontal and vertical sonar sensors. This system will form a 256 × 2740 pixel image containing depth information when scanning the front of the robot. In the process of moving forward, the positioning sensor will constantly update the position and attitude information of the robot in the global coordinate system, and save this information in the chip. The images detected by sonar will be solved in real-time, and the depth image information will be converted into the coordinate point information of the obstacles in front. If the coordinate point of the obstacle appears in the path originally planned in the global coordinate system, the automatic obstacle avoidance program will be triggered, so that the robot can bypass the obstacle.

Limitations and Future Outlook

Limitation

• Haven't taken battery endurance and watertight method into account, and it's difficult to select 'perfect' solutions that balance performance, development time and cost.
• Haven't communicated with related personnel or institution. A lack of practical experience may cause the project conception fails to be used in industry or research.

Future Outlook

• Short-term: Communicate with people to refine our ideas and realize each part of the functionality step-by-step
• Long-term: the 'plastivorous' fish may be used in various ways by carrying different sensors/actuators, and even more work as swarms to collect microplastics in the ocean (Become a part of the ecological environment!).

References

• [1] E. Guzzetti, A. Sureda, S. Tejada, and C. Faggio, “Microplastic in marine organism: Environmental and toxicological effects,” Environmental toxicology and pharmacology, vol. 64, pp. 164–171, 2018.
• [2] D. S. Green, L. Kregting, B. Boots, D. J. Blockley, P. Brickle, M. Da Costa, and Q. Crowley, “A comparison of sampling methods for seawater microplastics and a first report of the microplastic litter in coastal waters of ascension and falkland islands,” Marine pollution bulletin, vol. 137, pp. 695–701, 2018.
• [3] F. Xie, Q. Zuo, Q. Chen, H. Fang, K. He, R. Du, Y. Zhong, and Z. Li, “Designs of the biomimetic robotic fishes performing body and/or caudal fin (BCF) swimming locomotion: A review,” Journal of Intelligent &amp Robotic Systems, vol. 102, no. 1, apr 2021.
• [4] R. Wang, S. Wang, Y. Wang, L. Cheng, and M. Tan, “Development and motion control of biomimetic underwater robots: A survey,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 52, no. 2, pp. 833–844, 2022.
• [5] D. Scaradozzi, G. Palmieri, D. Costa, and A. Pinelli, “Bcf swimming locomotion for autonomous underwater robots: a review and a novel solution to improve control and efficiency,” Ocean Engineering, vol. 130, pp. 437–453, 2017.
• [6] D. Xu, Z. Lv, J. Liu, and J. Wang, “A novel artificial lateral line sensing system of robotic fish based on bp neural network,” in 2017 12th IEEE Conference on Industrial Electronics and Applications (ICIEA), 2017, pp. 1386–1390.
• [7] W. Shao and C. Xu, “Pitch motion control of a soft bionic robot fish based on centroid adjustment,” in 2019 IEEE International Conference on Mechatronics and Automation (ICMA), 2019, pp. 1883–1888.
• [8] K. Kanjanawanishkul, “Lqr and mpc controller design and comparison for a stationary self-balancing bicycle robot with a reaction wheel,” Kybernetika, vol. 51, pp. 173–191, 2015.
• [9] A. N. K. Nasir, M. A. Ahmad, and M. F. Rahmat, “Performance comparison between lqr and pid controllers for an inverted pendulum system,” AIP Conference Proceedings, vol. 1052, no. 1, pp. 124–128, 2008. [Online]. Available: https://aip.scitation.org/doi/abs/10.1063/1.3008655
• [10] . Braginsky and H. Guterman, “Obstacle avoidance approaches for autonomous underwater vehicle: Simulation and experimental results,” IEEE Journal of Oceanic Engineering, vol. 41, no. 4, pp. 882–892, 2016.
• [11] X. Cao, L. Ren, and C. Sun, “Research on obstacle detection and avoidance of autonomous underwater vehicle based on forward-looking sonar,” IEEE Transactions on Neural Networks and Learning Systems, pp. 1–11, 2022.
• [12] R. Katzschmann, J. DelPreto, R. MacCurdy, and D. Rus, “Exploration of underwater life with an acoustically controlled soft robotic fish,” Science Robotics,vol. 3, p. eaar3449, 03 2018.