Skip to main content

A Holistic Opto-Acoustic System for Monitoring Marine Biodiversities

Periodic Reporting for period 1 - SYMBIOSIS (A Holistic Opto-Acoustic System for Monitoring Marine Biodiversities)

Reporting period: 2017-11-01 to 2019-04-30

The development of fishing technology since the twentieth century has led to the emergence of fishing as one of the most serious problems for marine ecosystems. According to some estimates, if over-fishing is not brought under control, the entire fish stock of the world may collapse by 2048. Yet, decision making is based on defective existing methods. The process is subject to numerous sampling errors due to the limited statistics provided by such random and short-term sampling of the marine environment. Given these problems, there is a need to find solutions for real-time monitoring of target pelagic fish. The SYMBIOSIS project was chosen for the task and was granted funding to provide solutions to all the above-mentioned challenges and more.

SYMBIOSIS provides real-time monitoring of schools of fishes, including information about their size and movements in the deep sea, to help plan fishing policy in lakes, seas, and oceans and to facilitate the protection of the marine environment. The system is environmentally friendly, not only in terms of its own actions, which are non-invasive and do not influence the marine ecosystem, but also in the sense that it can provide reliable information about the condition of fish stocks in the sea. The SYMBIOSIS system operates on an entirely autonomous basis, and the sampling is undertaken over the long term and includes the capability to transmit information to a coastal center. In order to provide the authorities with concrete information, the research is focusing on the identification of six large fish species that are in very high demand by the fishing industry. As part of the project, a prototype is being developed including a system of acoustic sensors, a network of cameras, sophisticated processing units, and an energy unit permitting autonomous activity. The goal of the project is to sample the performances of the prototype system in three different marine environments: shallow Mediterranean, deep Mediterranean, and a tropical environment in the Canary Islands

To provide accurate detection and classification results, SYMBIOSIS includes a discovery chain that begins with the discovery and classification of fish, using cost-effective acoustic means. In order to conserve energy by activating the optical cameras only when a fish is present around the system, the acoustic system discovers and identifies the species of fish based on the typical speed and style of movement of each species. In addition, the acoustic sensors measure the size of the fish and the amount of fish in the school. After the acoustic system identifies one of the six selected species, it activates the optical system. This system includes sampling from several cameras and sophisticated data processing using a combination of algorithms for image identification and computerized learning. When the optical system confirms the identification of one of the six selected species, the information is transmitted to the surface by means of underwater acoustic communications, and then by radio communications to a coastal station.
All algorithms of SYMBIOSIS have been developed, implemented, and tested in numerical simulations and in multiple sea experiments. The system's hardware have been manufactured and tested per component in the lab and in a lake experiment, and deployment actions have been determined.

Objective 1:
Towards the aim of acoustic detection, we have developed two algorithms for coarse and accurate detection. We have tested the performance of both algorithms in multiple sea experiments and obtained 95% detection rate at false alarm of 0.001. For acoustic classification, we have explored two classification techniques - acoustic classification and optical classification, both tested in roughly 20 sea experiments involving ground truth information about the target. The performance showed correct classification of all 20 targets as fish with no false alarms. For the optical classification, so far we explored the performance for determining whether fish images belong to one of the six identified species with 95% accuracy. For biomass evaluation, we have developed an algorithm that measures the width of reflections identified as belonging to a fish that showed an error of less than 5% in the size of the mobile target in multiple sea trials. For data transmission, our results for image compression showed a compression rate of more than 50%, which yields files of roughly 5kB. The performance of the acoustic communication have been tested in a lake and showed communication availability of more than 95%. Finally, the RF link has been tested over the platform where the first Israel deployment of the SYMBIOSIS system will be carried out (namely, the THEMO buoy in Israel) and showed data rates of 30Mbits/sec with availability of more than 97%.

Objective 2:
We have completed the design of the SYMBIOSIS system and have implemented the hardware for it. A software package taking care of the interface with the hardware has also been set. The integrated system was tested in a lake experiment near Berlin, Germany, and proved the design concept of the system. In particular, the cameras identified a fish target with a very low distortion, and the acoustic system was able to detect a target deployed from a moving vessel in the lake. However, the algorithms have not been yet implemented on the actual hardware.

Objective 3:
While this activity is planned for the end of the project, we have already planned the actions for the deployment of the system on-board the offshore mooring across the shores of Israel. This activity will involve scuba divers disconnecting the main mooring cable, and connecting instead the system.
In terms of the expected impact of SYMBIOSIS, at this point, we can safely say that the project is meeting all its goals. In particular, SYMBIOSIS showed a high potential in acoustic and optic detection of fish. Different than current solutions that involves detection using a narrow beam or through anecdotal sightings and reports and thus cannot really monitor the entire area, SYMBIOSIS allows omni-directional detection up to a range of 500m. This enables the collection of data about the presence, composition (diversity), abundance, behaviour (e.g. migration) of the SYMBIOSIS key species of pelagic fish. Further, the SYMBIOSIS hardware was built for long-term autonomous monitoring. This enables enhanced knowledge regarding the movement and behaviour of fish, and their school formation. The project thus has the potential to become a standard in marine monitoring. The project has already led to commercial benefit in terms of the development of software defined USBL elements, optical cameras with extended detection range, acoustic detection of mobile targets, and acoustic localisation for wideband signals. The impact of the project is demonstrated by the identified 50 press releases identified for the project. SYMBIOSIS already made several advancements beyond the state-of-the-art. These activities have led to 11 peer-reviewed publications in key journals and conferences, with 5 more publications in a review process. And the work has been presented in 7 main conferences.
Testing of the system in a lake
Example of acoustic detection
Example of detection of mobile targets
Example of optical detection
Example of image compression
A scheme of the symbiosis unit
Block diagram of the SYMBIOSIS process
Example of acoustic localisation results