Motivation

Climate change is already affecting living conditions in marine ecosystems significantly. Especially, commercially important fish species are endangered by the increase of heatwaves, which are leading to dead zones in shallow coastal waters. Preliminary work suggests focused monitoring of fish swarms in coastal regions, e.g. in the western Baltic Sea, to gain a deeper understanding of the marine ecosystem dynamics and its response to climate changes. However, those regions cannot be accessed by standard research vessels. In addition, the traditional surveys to determine the state of fish stocks are usually based on net catches and are therefore invasive, i.e. large biomasses are often taken from species that should be protected. Another major drawback is that the existing methods only provide snapshots of the state of fish stocks with low spatial and temporal resolution.

Goals and Contributions

In cooperation between three Universities of Hamburg (HAW, TUHH, UHH) an interdisciplinary research team has been formed to investigate the behavior and migration patterns of endangered fish populations. The goal of the BeoFisch project is to develop a novel method for non-invasive, high spatial, and temporal resolution monitoring of the state of fish stocks in coastal regions. Figure 1 illustrates the proposed solution concept. During the project, machine learning (ML) algorithms will be developed for the detection and classifications of fish swarms, which will then run on autonomous underwater vehicles (AUVs) to track the fish populations. Prototypes based on the commercial BlueROV [1] will be developed at the smartPORT group (TUHH) and equipped with the in-house built acoustic modem (ahoi) to establish a mobile underwater communication network. This allows not only the live monitoring of fish swarms but also plays an important role in the localization of the AUVs. The main challenges associated with acoustic localization are the low bandwidth and the low update rates, which results in a delay of several seconds between distance measurements. For this purpose, the smartPORT group will develop and implement suited sensor data fusion and control algorithms to realize the robust navigation and coordination of multiple networked AUVs.

References

[1]: Blue Robotics Inc. (2020, April 26). BlueROV2 - Affordable and Capable Underwater ROV. [Online].
Available: https://bluerobotics.com/store/rov/bluerov2/