Using blockchain and AI technologies for sustainable, biodiverse, and transparent fisheries of the future

Abstract

This paper proposes a total fusion of blockchain and AI tech for tomorrow’s viable, rich in diversity and transparent fisheries. It outlines the main goal of tackling overfishing challenges due to lack of transparency and biodiversity depletion in the fisheries sector. With the use of blockchain technology, we can ensure that all fishery products are safely traced from their harvest up to when they get to the market— at the same time, AI algorithms are used in monitoring fish populations and predicting them plus decision-making processes which should be enhanced thus promoting bio-diversity and ensuring sustainability of fish stocks. Results show promise on using both technologies together: improving sustainability plus transparency in fisheries which would promote more fish biodiversity, while others including using an artificial intelligence system have not been confirmed yet by observations. The conclusion underscores the transformative nature of these technologies as having great implications towards fisheries management; this implies that there is a need for future observational studies aimed at validating such other findings.

Introduction

The exploitation of marine resources can have detrimental consequences for the environment. The implications of these impacts on ecosystems are of great importance and should influence our perception of ‘acceptable’ levels of marine resource use and associated impacts. It is crucial to recognize that the risks associated with this exploitation include many factors, in addition to the direct effects on the target species. The species most vulnerable to the consequences of fishing are likely to be those with low reproductive capacity, which can take decades to reproduce. Marine fishing faces a number of challenges, including: [1, 2].

• Overfishing is one of the biggest challenges facing marine fisheries. It occurs when fish catch rates exceed the reproductive capacity of populations, leading to the depletion of fish stocks. Fishing activities, particularly certain methods such as bottom trawling, can cause environmental damage, including the destruction of marine habitats, disruption of ecosystems, and generation of plastic waste. Water pollution from waste, chemicals, and hydrocarbons originating from fishing boats, transport vessels, oil tankers, and even tourist boats can negatively impact water quality and the health of marine ecosystems [3].

• Inadequate fisheries management: Ineffective fisheries management, including lax regulations can lead to imbalances in fish populations and conflicts between fishermen. - Impacts of climate change: Climate change affects the oceans by altering sea currents, water temperature and the distribution of marine species, which can disrupt fisheries. Loss of biodiversity: Fishing can lead to by-catches, including of non-targeted or endangered species, which can contribute to the loss of marine biodiversity [4, 5].

• Socio-economic issues: Fishing communities depend on marine resources for their livelihoods. Economic challenges, competition and the equitable distribution of the benefits of fishing are major concerns [6].

• Labeling and traceability: Accurate identification of the origin of fish products and their traceability throughout the supply chain are crucial issues in the fight against illegal, unreported and unregulated fishing. [7, 8]. As traceability becomes increasingly important, organizations are expanding their services and programs. However, companies have diverse traceability needs based on their size, location, product types, and position within supply chains. Companies at the end of the supply chain, like retailers and consumer-facing businesses, face challenges in managing large volumes and diverse product ranges. Near the point of harvest, harvesters struggle to collect and store reliable data due to harsh environments and require durable traceability tools. Aquaculture facilities must monitor and share information about their inputs and other data down the supply chain. Mid-supply-chain companies, such as processors, need traceability systems to track product transformations like weight, composition, and packaging changes. Upstream businesses use traceability to support product quality and integrity claims, while consumer-side businesses leverage it to educate customers and ensure compliance with due-diligence policies.

In order to address some of these challenges, the integration of Internet of Things (IoT), artificial intelligence (AI), and Blockchain technologies is emerging as a promising solution. These technologies come together to create a smart fisheries system, which facilitates real-time monitoring of fishing operations, efficient management of marine resources, and transparent tracking of the supply chain [3]. Given this perspective, this article aims to delve into the essential role played by IoT, AI, and Blockchain in the establishment of smart fisheries with a specific focus on traceability. We will present the theoretical underpinnings of these technologies, examine their application within the fisheries context, and emphasize the individual advantages they bring to ensure the sustainability of fishing activities.

The primary objective of this article is to provide a comprehensive exploration of the potential offered by the combination of IoT, AI, and Blockchain to enhance traceability in the fishing industry. Our paper consists of six parts. We begin with a state-of-the-art review on the application of Blockchain in the environmental field, particularly focusing on biodiverse and transparent fisheries of the future. Following this, we detail not only Blockchain technology but also IoT In aquaculture in Sect. 3. Section 4 proposes our implemented approach and application. The results of our work are revealed in Sect. 5. Finally, Sect. 6 is divided into three parts: Discussion, Recapitulation, and Future Work, providing a thorough conclusion and perspectives for future research and development.

Related work

Blockchain emerges as a disruptive technology offering a solution to instill trust among skeptical parties [9]. At its core, Blockchain comprises a distributed, decentralized, and inherently connected data structure known as a ledger [10]. Implementing Blockchain necessitates a peer-to-peer network of independent peers, where data is stored in cryptographically linked blocks and replicated across each peer, ensuring that every participant maintains a consistent local copy of the ledger. A comprehensive Blockchain platform supports both data and metadata storage, allowing data to trigger self-executing workflows called smart contracts. The cryptographic guidelines and consensus protocols embedded within Blockchain ensure data integrity and consistency [11, 12].

Consensus protocols stand out as one of the critical aspects influencing Blockchain performance. Existing literature broadly categorizes consensus into two main groups: evidence-based and vote-based. Evidence-based consensus mechanisms (e.g., PoW, PoS, PoET, PBFT, PoI) typically operate akin to a lottery, enabling any peer to participate without prior authorization [13].

Beyond cryptocurrency, Blockchain finds applications across various domains such as healthcare, finance, e-voting, the energy sector, gaming, etc., aiming to enhance reliability, transparency, and security. We outline diverse existing Blockchain-based solutions tailored for the traceability of fishery products.

Smart fisheries have gained traction in the global seafood market in the 21st century as a means to address the increasing demand for seafood while mitigating declines in fish populations. One notable application of Blockchain in fisheries involves efforts to bolster tuna traceability around Fiji and other South Pacific islands [14, 15]. To combat illegal, unreported, and unregulated fishing, participating fisheries tag fish with either radio frequency identification (RFID) or quick response (QR) tags immediately after capture. These tags are then automatically recorded at various checkpoints in the supply chain, with each record logged into the Blockchain. For instance, in the South Pacific tuna fishery, a UK-based company “Provenance” has established a Blockchain platform for tracking agricultural, forestry, and fisheries products, complete with fair and organic consumption labeling.

Niya [14] presents a distributed system leveraging IoT and Blockchain for automatic measurement, storage, and monitoring of water and air quality in diverse environments such as lakes, mountains, regions, cities, or factories. Such advanced solutions combine human interaction for data access, demand high power or space requirements, or are based on centralized architectures. The pollution monitoring system offered by Niya utilizes LoRa to address IoT protocol challenges related to high energy consumption and long-distance transmission, while also ensuring full decentralization through the Ethereum Blockchain for data storage and retrieval recorded by IoT sensors. Consequently, data integrity is maintained without the need for a trusted third party, and data collection and capture are automated with minimal manual effort. Observations from four different sensor types measuring pH, turbidity, carbon monoxide (CO), and carbon dioxide (CO₂) demonstrate high accuracy within expected measurement times and no falsification of collected experimental values, serving as reliable evidence of contamination presence.

The SMARTFISH H2020 project [15] proposes a system to optimize fisheries ecosystem efficiency while mitigating the ecological impact of the EU fisheries sector. Leveraging cutting-edge technologies such as big data, deep learning, and artificial intelligence, the project aims to assist fishermen in making informed decisions. IoT devices deployed within the fishing industry interconnect to gather data from sensors measuring pH, temperature, salinity, etc., empowering fishermen with IoT applications. Over the next three decades, the fisheries ecosystem alone could generate $60 billion in annual benefits through advanced analytics [16].

A solution, developed by [17], employs the Ethereum Blockchain to secure data in maritime surveillance systems. Using buoys and drones to collect detection data at sea, the system transmits this data via a mesh topology network to a ground-based data fusion center. The authors propose a proof of authority consensus protocol, which is more energy-efficient and tailored for authorized Blockchains. MATLAB simulations evaluate the system’s performance in terms of latency and throughput, revealing a manageable latency overload of 13–18% and throughput of 12–16%, meeting acceptable quality service levels.

A report published by the Food and Agriculture Organization of the United Nations [18] explores various ways to harness Blockchain technology in the seafood supply chain. To address existing challenges in the seafood lifecycle, researchers examine the application of Blockchain in seafood supply chain management. They suggest that integrating NFC, RFID, and QR code techniques could help combat traceability issues. However, the study does not present a prototype and focuses solely on seafood supply chains, overlooking farmed fish. Consequently, species substitution may occur due to externally caught fish introduced into the ecosystem. The study also identifies unresolved issues in the seafood supply chain, such as fish label and tag damage during transportation.

Authors in [19, 20] investigate how Blockchain technology can enhance global ocean conservation and fisheries supply chain management. Blockchain finds increasing utility across a spectrum of services and industries, including transparent resources for ocean conservation, plastic pollution reduction, slavery at sea mitigation, and sustainable fisheries management. Public skepticism regarding certain conservation efforts and seafood provenance is on the rise. While some global marine conservation groups and seafood producers leverage disruptive technologies like Blockchain, meaningful participation from coastal communities and artisanal fishermen is necessary to harness benefits equitably.

Freire et al. [21] present an authorized Blockchain solution prototype using Hyperledger Fabric, a robust, modular and efficient open source Blockchain platform. They evaluate the performance of the solution through practical experience in which a prototype receives detection data from a receiver of the automatic identification system based on a low-cost software-defined radio built using a Raspberry Pi. In order to reduce scalability attrition, they developed a Blockchain client that was delivered for easy large-scale deployment. In addition, the researchers determined the optimal hardware configuration for their customers through extensive experimentation, while aiming to reduce implementation and maintenance costs. Performance results quantify the overhead costs of Blockchain technology and its impact on quality of service.

The main findings of this review indicate that Blockchain can play a crucial role in combating overfishing by providing transparent traceability of the supply chain from capture to commercialization. Through the use of IoT sensors and tracking systems, data on the origin and quality of fishery products can be immutable and accessible in real time. In addition, Blockchain offers the opportunity to set up smart contracts to facilitate agreements between fishing stakeholders, including fishermen, processors, distributors and regulators. These smart contracts can automate payments and licenses, reducing bureaucracy and improving operational efficiency.

Theoretical foundations

This section explores the theoretical underpinnings of IoT and its application in the field of aquaculture, highlighting both common and specific challenges faced in the industry.

Challenges of IoT in Aquaculture

Modern fisheries depend heavily on the IoT, which offers real-time data and automation capabilities that increase productivity and efficiency. Nonetheless, there are particular difficulties when integrating IoT in aquaculture [22].

Common challenges

One prevalent challenge involves the continual contact of sensors with water, which often leads to reduced accuracy over time due to contamination. Solutions proposed to address this issue include regular manual cleaning of probes or the implementation of automated cleaning mechanisms. The overall cost of IoT solutions is increased by such mechanisms, but accurate readings are ensured and the need for manual maintenance is minimized. Effective automation of the cleaning process has been explored through innovative approaches, including the use of air bubbles or ultrasonic cleaners [23].

The importance of data accuracy and reliability in aquaculture IoT systems cannot be overstated when it comes to making informed decisions. Data quality can be compromised due to device errors and network issues, which then require maintenance procedures and the integration of data fusion techniques and anomaly detection algorithms. Furthermore, network-related challenges can be mitigated by optimizing network configurations and implementing topological designs that improve data transmission.

Limited access to electricity and internet connectivity are among the challenges related to infrastructure. Innovative solutions are needed due to the lack of reliable electricity in rural areas or marine environments, as well as the unreliability of urban electricity. Despite being designed for low power consumption, IoT devices still face persistent challenges, which necessitate careful consideration of alternative power sources like solar panels and photovoltaic batteries. Energy optimization strategies, including reducing data transmissions and implementing sleep modes, further extend battery life. Addressing internet accessibility issues involves exploring different networking protocols and leveraging advancements such as edge computing to reduce reliance on cloud resources [5, 23].

Specific challenges

Specific challenges in the integration of IoT technology in aquaculture vary based on farm accessibility and species sensitivity. Farms with limited access to basic facilities like electricity or internet face more significant hurdles, especially when dealing with sensitive species that require higher reliability and quality in design and technology. Categorizing challenges based on farm conditions reveals distinct scenarios, each with its own set of infrastructure, data, and perception-related challenges.

Access to electricity and internet connectivity poses significant infrastructural challenges for IoT systems in aquaculture. The lack of reliable electricity in rural areas or marine environments, alongside the unreliability of urban electricity, necessitates innovative solutions. IoT devices, although designed for low power consumption, still face challenges that require careful consideration of alternative power sources, such as solar panels and photovoltaic batteries. Energy optimization strategies, including reducing data transmissions and implementing sleep modes, are essential to extend battery life. To address internet accessibility issues, exploring different networking protocols and leveraging advancements such as edge computing can reduce reliance on cloud resources.

Ensuring data accuracy and reliability is critical for informed decision-making in aquaculture IoT systems. Device errors and network issues can compromise data quality, necessitating maintenance procedures and the incorporation of data fusion techniques and anomaly detection algorithms. Additionally, optimizing network configurations and deploying topological designs that enhance data transmission can mitigate network-related challenges.

Perception barriers, particularly among small-scale farmers, pose challenges to IoT adoption in aquaculture. The lack of awareness and economic constraints hinder adoption, necessitating initiatives such as pilot programs and economic incentives to demonstrate the effectiveness and affordability of IoT solutions. Overcoming perceptual barriers requires proactive engagement with farmers and governmental support to promote technology adoption and improve confidence in technological advancements. In addition, financial help and cost-effective solution exploration can enable small firms to use IoT technology [23].

Through the resolution of these particular obstacles, the integration of IoT in aquaculture may be maximized to guarantee dependable and effective operations, thereby augmenting the industry’s sustainability and production.

Smart fisheries and A.I

AI holds the promise of revolutionizing the fishing industry, enhancing operational efficiency, and promoting resource conservation. By analyzing vast amounts of data collected through the IoT, AI offers valuable insights to optimize fishing operations.

Five sensors “(a dissolved oxygen sensor (DO sensor), temperature sensor, turbidity sensor, pH sensor, and a proximity switch sensor used for location identification purposes)” and five actuators (a heater, a C⋅W⋅/ C⋅C.W. motor used for wind-proofing purposes, a water pump, an agitator, and a feeding device driven by servo motor) have been integrated into the intelligent fish-feeding system. A mobile application allows remote monitoring of water quality parameters measured by the smart pond via a cloud server. Two modes (auto mode and manual mode) can be selected in the interface of the cell phone. IPCAM is installed onto the fish pond to monitor fish baiting status and record fish length. An alarm system notifies users when a parameter for the farm exceeds a specific critical threshold. The system controls water quality parameters (dissolved oxygen level, water temperature value, turbidity level, and pH value) to the targeted region. Environmental data and site photos stored on the cloud serve as a portfolio for potential customers. A deep learning model predicts the outcome of the developed smart fish pond (California Bass) with different variable parameters. Statistical error reduction methods improve prediction accuracy. The model precisely anticipates the output properties and demonstrates viability and applicability under different input parameters. Real-time farm information aids in identifying problems before they occur. Predictive models may provide farmers with dissolved oxygen concentration value a day in advance for proactive adjustments. The SHAP library is used to depict the impact of each input parameter on the output of the generated optimal model [24]. Chiu et al. [25] propose an approach to address these challenges, focusing on real-time monitoring and control to optimize feeding practices for long-term sustainability and reduce environmental impact. The results can significantly influence on-farm management by optimizing feeding for long-term sustainability and reducing environmental impact. An innovative smart aquaculture farming system will be developed to enhance feed utilization efficiency and provide tailored production recommendations by integrating water monitoring data with production and meteorological data.

Utilizing machine learning algorithms, AI has the capability to forecast seasonal patterns in fish populations [6], precisely recognize species, and even detect illicit fishing practices. Through route optimization, AI enables fishermen to minimize unnecessary journeys, thereby lowering their environmental impact while maximizing their catch. Moreover, AI is instrumental in mitigating non-target bycatch by identifying unintended species caught inadvertently, empowering fishermen to take measures to mitigate adverse effects on marine ecosystems.

In [23], Ezzedini et al. study presents a comparative examination of Faster R-CNN and YOLOv8 for real-time identification of fishing vessels and fish in maritime surveillance. It emphasizes the importance of this research in advancing fisheries monitoring and object detection through deep learning techniques. By specifically evaluating the performance of Faster R-CNN and YOLOv8, the study aims to clarify their effectiveness in real-time detection, emphasizing the significance of such capabilities in fisheries management. Through a comprehensive literature review, the research establishes the current state-of-the-art in object detection, particularly in the realm of fisheries monitoring, while addressing existing methodologies, challenges, and constraints. The results not only demonstrate the superior precision of YOLOv8 in detection but also underscore its potential impact on maritime surveillance and the preservation of marine resources.

In addition, detection of underwater objects presents itself as a major challenge in this field, due to the complex and changing conditions of the underwater environment. Limited visibility, lighting variations, and water turbidity make this task complex, requiring innovative solutions to ensure effective and reliable monitoring. It is in this context that the use of the YOLOv8 algorithm proves promising [26, 27]. At the same time, data security and transparency in the collection and processing of underwater information are critical concerns.

Introduction to Blockchain technology and its benefits for traceability

Blockchain technology is a decentralized, secure and transparent registry that records all transactions in a chronological and immutable manner [28]. In the fisheries context, Blockchain offers a reliable solution to ensure end-to-end traceability of the seafood supply chain.

The utilization of Blockchain technology enables the comprehensive and immutable recording of every stage of the fishery process, spanning from capture to distribution. Critical details including date, time, location of capture, species caught, fishing methods employed, and vessel information are securely recorded within interconnected blocks [6].

This heightened transparency afforded by Blockchain empowers consumers, regulators, and stakeholders to verify the authenticity and origin of seafood products [6]. By leveraging Blockchain, the fishing industry mitigates the risks associated with data falsification, fraud, and illegal fishing activities, thereby fostering trust and advancing sustainability and responsibility within the sector.

Blockchain integration for traceability

Integrating blockchain technology into smart fisheries provides numerous advantages that enhance the security, transparency, and efficiency of the entire supply chain. This section explores the various benefits and applications of blockchain in this field.

Advantages of blockchain for smart fisheries

The incorporation of Blockchain technology into the realm of smart fisheries presents numerous notable benefits. Foremost among these advantages is the security and reliability inherent in this technology. Utilizing sophisticated cryptography mechanisms, Blockchain ensures the integrity of data stored within blocks, rendering it virtually immune to tampering or falsification [29]. This secures the information’s integrity throughout the entire supply chain, from the moment fish are caught to their eventual distribution.

Another key advantage lies in the decentralization aspect of Blockchain. Unlike centralized systems where data resides on a single server, Blockchain operates across a network of distributed computers (nodes) spanning the globe. This eliminates single points of failure and enhances the system’s resilience against cyber-attacks or potential errors [30]. Consequently, smart fisheries reap the benefits of heightened robustness and increased data availability.

Creation of a decentralized fish catch Registry

Blockchain facilitates the establishment of a decentralized register of fish catches, meticulously documenting all transactions and activities associated with fishing in a transparent and immutable manner. Every time a fish is captured, crucial details such as species, size, fishing method, date, and location of capture are meticulously recorded within a Blockchain block. These blocks are then interconnected chronologically, forming an unbroken chain of custody [31].

This decentralized registry offers seamless traceability, enabling each fish to be traced from its initial point of capture to its sale to the ultimate consumer. Consumers can thus access comprehensive information regarding the origins of the seafood they purchase, bolstering their confidence in the quality and sustainability of the products.

Enhanced transparency and trust in the supply chain

Blockchain enhances transparency within the smart fisheries supply chain. By accessing immutable Blockchain data, stakeholders including consumers, governments, NGOs, and businesses can scrutinize the entire fishing process, from capture to sale, without the risk of data manipulation [18].

This transparency fosters trust among industry participants and promotes ethical business practices. When consumers have assurance that seafood originates from legal and sustainable sources, they are more inclined to support environmentally responsible fisheries. Companies can leverage Blockchain traceability as a competitive edge by showcasing their dedication to sustainability and environmental stewardship.

In conclusion, the integration of Blockchain into smart fisheries offers substantial benefits, including data security, decentralization, comprehensive traceability of fish catches, and transparency in the supply chain. These collective advantages contribute to establishing a sustainable and transparent fishery, thereby safeguarding marine ecosystems and meeting the increasing demands of consumers concerned about the provenance and sustainability of seafood products.

Proposed approach

This platform, which builds upon several previous smaller projects, integrates various established components to create a comprehensive solution. While it leverages existing technologies, it combines them in a novel manner to enhance functionality and user experience. Crafted with the aim of empowering fishermen, it not only facilitates the gathering of precise information regarding marine resources and environmental conditions but also streamlines the process of sharing this invaluable data while navigating the open waters. By securely transmitting the collected data to the Marine Office, this platform ensures that vital information is stored in a highly secure and easily accessible manner.

Equipped with state-of-the-art wireless sensors boasting advanced capabilities, fishermen are equipped to capture a wealth of crucial details including precise GPS coordinates, comprehensive catch volumes, accurate species identification, nuanced salinity and pH levels, as well as detailed weather conditions. Through instantaneous transmission of this wealth of data to the marine office, stakeholders gain immediate insights into the dynamic ecosystem of the seas.

Leveraging the power of decentralized Blockchain technology, the platform employs robust encryption, meticulous time-stamping, and impregnable secure storage mechanisms to safeguard against any potential manipulation or tampering of the collected data. This ensures the integrity and reliability of the dataset, empowering marine authorities to harness advanced analytical tools to discern emerging trends, conduct comprehensive assessments of fish stocks, and implement proactive measures for the sustainable management of marine resources. (Fig. 1)

Fig. 1

Smart fish platform based on blockchain

Moreover, the platform serves as a conduit for seamless collaboration between fishermen and regulatory bodies, facilitating the exchange of critical insights and fostering an environment conducive to the enhancement of fisheries management policies, preservation of delicate marine ecosystems, and promotion of sustainable fishing practices. By revolutionizing the landscape of fishing data collection and management, this visionary platform endeavors to safeguard marine resources for future generations, champion sustainable fishing endeavors, and cultivate a culture of transparency and trust among all stakeholders involved.

To further enrich the platform’s functionality, we’ve integrated an advanced Faster R-CNN (Region Based Convolutional Neural Networks) model for image analysis of the fish captured by fishermen [32]. This state-of-the-art deep learning object detection system automatically identifies the species of fish and accurately estimates their sizes. By simply uploading images of their catches, fishermen can leverage this cutting-edge technology to gain valuable insights into their hauls, without requiring specialized knowledge of species identification.

The project harnesses the power of a specialized Faster R-CNN model as its primary object detection algorithm for fish identification. By leveraging this optimized model, it meticulously analyzes images of fish captured by fishermen, precisely discerning fish contours and extracting detailed information about the species comprising the catch. Following detection, the pertinent data, such as identified species and corresponding GPS locations, is securely recorded within a decentralized Blockchain database. This seamless integration not only ensures a streamlined and transparent process but also enables more accurate and sustainable management of marine resources, thereby enhancing the platform’s efficacy in promoting responsible fishing practices and ecological conservation.

The Faster R-CNN [32] model proved to be the optimal choice for this application, offering high performance with an execution speed of 82 frames per second, and a high detection accuracy of 98,67%, surpassing other object detection algorithms such as SSD (Single Shot Detector) and RCNN [23]. This combination of wireless sensors, object detection by the Faster R-CNN model and secure storage by the Blockchain puts our application at the forefront of technology for a smart, sustainable and responsible fishery.

Methodology & results

Implementation

The work proposed in this system applies to an important need in the process of protecting marine fauna and flora. Indeed, there have been continuous complaints and arrests of maritime fishermen due to their negligence and non-compliance with maritime legislation. This is due to the use of traditional and archaic fishing methods in the Tunisian context.

As part of this project, we aim to produce a proof of concept (PoC) to optimize the fishing methodology and determine the locations of fish groups, using marine cameras. A GPS location will be sent to the various fishermen to help them locate the fish and thus minimize energy consumption and maximize the result of legal fishing. The platform’s sensors will be the marine camera and the geolocation sensor. These data are collected by a Raspberry PI 4 platform and then sent. The images will be saved on a database, and transactions will be sent via Web 3.0 to be saved on an Ethereum blockchain. Figure 2 illustrates this PoC.

Fig. 2

Proposed application description

In our project, we harness Ethereum Blockchain technology, complemented by Truffle and Ganache tools, to streamline the creation and implementation of smart contracts. These contracts, scripted in Solidity, Ethereum’s dedicated coding language, offer extensive functionalities supported by Ethereum’s dynamic open-source community. With digital signatures, the integrity of smart contract source code remains intact, ensuring the safety and dependability of transactions. Every activity conducted on the platform is considered a transaction, mandating the payment of gas fees in Ether, Ethereum’s native digital currency, to facilitate seamless processing.

Our approach is based on an authorized Blockchain network, where only valid and authorized nodes can connect and participate in the network. All miners involved in the Blockchain network must have adequate performance and memory capacity for mining, a mathematically intensive task to secure the network. By using Ethereum as a Blockchain infrastructure, we ensure that smart contracts are irreversible and non-modifiable once integrated into the system, ensuring the integrity and transparency of captured data and messages transmitted. The aim of this platform is to have a totally autonomous system. Each fisherman who joins the platform will have access to its benefits, and will consequently be an autonomous, automated node of the Blockchain, able to record data directly on the Blockchain.

The trio of smart contracts, namely FishersContract, MarineOfficeContract, and RescueAgentsContract, are intricately crafted to collaborate and enable an intelligent platform for the real-time collection of fishing data. They integrate seamlessly with a decentralized Ethereum Blockchain to ensure the security and transparency of the data.

Figure 3 illustrates the contract executed by fishermen account to capture and submit their fishing data, such as GPS location, humidity, salinity, pH, oxygen level, chlorophyll, as well as the existence of fish and possible emergency alerts.

Fig. 3

Fishers contract

This contract also ensures the registration of fishermen and provides mechanisms to ensure that only registered fishermen can submit their data (See Fig. 4).

Fig. 4

Is the marine office contract

It is executed to receive, store and manage fishing data transmitted by fishermen.

This contract allows the Office Marine to consult and verify the information submitted by each fisherman, thus contributing to the sustainable management of marine resources. The Marine Office can also send emergency alerts to fishermen in case of need.

Figure 5 RescueAgentsContract is the contract executed to receive emergency alerts issued by the Marine Office or other fishermen. Emergency Wardens can register on this contract and receive alerts with information about the sender of the alert and the emergency message.

Fig. 5

Rescue agent contract

The relationship between the contracts is such that fishermen use FishersContract to submit their fishing data, which is then received and managed by the Marine Office through MarineOfficeContract. In the event of an emergency alert, the Marine Office can send these alerts to the Emergency Workers via RescueAgentsContract, allowing a rapid response in the event of a critical situation at sea. For the purposes of this paper, we have not described all the functions implemented. However, each fisherman, by registering and joining the system, will be automatically added and will have access to the various functions relating to each member of the node. There is no need for human intervention, as the system is fully automated, so when the boat starts up, everything is triggered automatically.

As mentioned, we worked on a proof of concept As a result, we deployed our system using remix for smart contracts, Ganache for transaction testing and matamask for payment of the various transactions.

Ganache results

Ganache stands as a favored development tool for crafting and simulating a local Ethereum Blockchain [33, 34]. Its user-friendly graphical interface enables real-time tracking and debugging of transactions, events, account addresses, and balances, simplifying interactions with smart contracts. Three contracts have been deployed on Ganache to establish the foundation of the real-time fishing data collection platform, known as SmartFish.

FishersContract

• Contract Address:

0 × 31605a4835a01536a79e98685e1b01c3cf5fc4cb.

Description: This contract allows fishers to capture and submit real-time fishing data such as GPS location, humidity, salinity, pH, oxygen level, chlorophyll, as well as information on the existence of fish and emergency alerts. Fishers can register, submit their data and send possible alerts.

MarineOfficeContract

• Contract Address:

0 × 3753e7d89237aacd26e088b61f4f28812c0bdd32.

Description: This contract is used by the Marine Office to receive, store and manage fishing data submitted by fishers. It can also receive emergency alerts from fishers. The Marine Office is responsible for the sustainable management of marine resources.

RescueAgentsContract

• Contract Address:

0 × 5e7c0091a9efc8b54407c079a750e672ef7378b5.

Description: This contract allows Emergency Wardens to register and receive emergency alerts issued by the Marine Board or other fishers. Emergency Wardens can respond quickly to critical situations at sea. (Figures 6 and 7).

Fig. 6

Deployment of smart contracts

Fig. 7

Transactions in fisherscontract

For FishersContract contract transactions: (Figures 8 and 9)

• Hash Transaction:

0xd0af92b5548a8ffded98ae5ac729884f4ab354145030b1c8b96979016161a0d3.

• Sender address:

0 × 7207904b4fb8471ba1bcfda82a7c7c5a8dd5cd0b.

• Target contract address:

0 × 31605a4835a01536a79e98685e1b01c3cf5fc4cb (FishersContract).

• Transaction Data (TX DATA):

• Called function: captureData (gpsLocation: string, humidity: uint256, salinity: uint256, pH: uint256, oxygenLevel: uint256, chlorophyll: uint256, fishExistence: bool).

• Inputs:

• gpsLocation: “N 34° 42’ 58.9248, E 10° 45’ 55.17”.

• humidity: 55.

• salinity: 38.

• pH: 8.

• oxygen level: 7.

• chlorophyll: 10.

• fishExistence: true.

Fig. 8

Events in fisherscontract

FishersContract Contract Events:

• Event issued: DataCaptured.

• Fisherman’s address:

0 × 0b35b01225221533cf73b71133277f6517243c83.

• GPS location (gpsLocation: “N 34° 42’ 58.9248, E 10° 45’ 55.17”.

Fig. 9

Add block in blockchain

Block of FishersContract:

• Block number: 29.

• Used gas limit: 117,610.

• Block timestamp: 2024-07-10 01 :17 :23.

• Block hash:

0 × 507c64e824e923521ff20168cd8b204173a0e99c6bab114627b4ae1c2c5f19ab.

Yolo results

As a second solution, we employed also the YOLOv8 model on Google Colab for object detection in videos depicting various marine scenes. Videos were converted to images, and a caption ontology was defined to categorize objects of interest such as fish, algae, corals, shipwrecks, sharks, sea turtles, shells, seabirds, plastic waste, trash oil/petroleum, and non-plastic waste. Image labeling was performed with autodistill, contributing to the creation of an annotated dataset for training. The experiment took place on a Google Colab environment, using an Intel Core i5 processor, an NVIDIA GeForce MX350 GPU, and 8 GB of memory. Expected results include accurate detection of marine objects, with performance evaluations based on metrics such as precision, recall, and F1 score. Considerations were made regarding training data quality, hardware specifications, and the need for parameter adjustments to optimize results.

Figure 10 comprises four parts: Fig. 10a presents the detection results on the dataset using the YOLOv8 model. It illustrates how different classes of marine objects, such as fish, algae, and corals, are identified and classified in various images. Figure 10b shows further detection results from the same dataset, highlighting the model’s ability to recognize more marine objects, including shipwrecks and sharks, providing a comprehensive view of the detection accuracy across different categories. Figure 10c continues to demonstrate detection results, emphasizing the identification of objects like sea turtles, shells, and seabirds. It showcases the model’s versatility in detecting diverse marine life and environmental elements. Figure 10d displays additional detection results, focusing on plastic waste, oil/petroleum, and non-plastic waste in the marine environment. This part underscores the model’s effectiveness in detecting pollution and other non-living objects, which is crucial for environmental monitoring and conservation efforts.

Fig. 10

(a). Detection results on dataset. (b). Detection results on dataset. (c). Detection results on dataset. (d). Detection results on dataset

The results of the evaluation of the YOLOv8 model on the classes defined in the caption ontology show a solid overall performance, although variations are observed between different categories of marine objects. The main metrics evaluated include precision (P), recall (R), mAP50 score (average precision at a 50% confidence threshold), and mAP50-95 score (average precision over a range of confidence thresholds of 50–95%). (Table 1)

Table 1 Yolov8 detection results

Overall, across classes, the model achieved an mAP50 score of 0.621, indicating a reasonable ability to locate and classify objects of interest in images. The “Shells” class showed the best performance with a mAP50 of 0.735, closely followed by “Algae” with a score of 0.720. On the other hand, “Seabirds” recorded a lower mAP50 of 0.574.

Discussion, recapitulation and conclusion

In this section we are discussing the outcomes of the project.

Discussion

Incorporating Blockchain technology into the SmartFish platform yields promising outcomes for the fishing sector. Through the deployment of three smart contracts on Ganache, a robust platform has been established, empowering fishermen to capture and promptly submit critical marine and environmental data in real-time. Leveraging Ganache, a local Blockchain, streamlines the development and testing of contracts, ensuring their efficiency and reliability before deployment on the Ethereum Blockchain in live environments.

The fusion of AI plus IoT with Blockchain and its significance mainly lies in revolutionizing — through real time data collection, analysis and management — the processes. The ability to collect data in real time has helped fishermen identify best fishing grounds and detect possible threats (while preserving marine ecosystems) through use of IoT sensor devices. On the other hand, detection of emerging trends through AI analysis on large datasets helps in accurate assessment of fish stocks; this assists in coming up with proactive measures for sustainable resource management that fosters resilience as well as responsibility in the fishing industry. However, there are considerations around data confidentiality and ownership. Fishermen, governments, and private groups involved in the adoption of this technology need to agree on how data will be collected, shared, and used. Blockchain technology allows hashing (SHA256, Keccak) and encryption of data with asymmetric keys (public/private keys), ensuring the security of private data, especially by using digital certificates in each connected fishing port.

The cost of adopting Blockchain technology in fisheries (estimated at an additional 5–10% of the cost) can be borne by a combination of parties involved, including fishermen, governments, and private groups. The specific distribution of costs will depend on various factors such as the scale of implementation, turnover, regulations, and policies in place.

• Fishermen may bear some of the costs associated with the adoption of Blockchain technology, particularly if this involves them jointly purchasing new equipment.

• Governments can play an important role in funding or subsidizing the adoption of Blockchain technology in the fisheries sector. This can be done through grants, subsidies, or other financial incentives to encourage fishermen to adopt the technology.

• Seafood companies can also contribute to the costs of adopting Blockchain technology. These groups may see value in improving traceability and sustainability in the seafood supply chain and may be willing to invest in technologies that support consumer confidence. The outcome is the ability to assure customers of safe products and provide guarantees on product provenance.

To ensure fishermen’s data is not compromised by security breaches while also keeping their privacy intact, it might be worth looking at Hyperledger Aries. This is a system of decentralized identity management infrastructure — an innovative choice for future implementations. On the other hand, merging with YOLOv8 brings in the capabilities of a deep learning model to detect objects. This way, our system can automatically recognize fish shown in pictures or clips— an action point that lightens the load on fishermen and at the same time streamlines data collection efficiency.

The deployment of Blockchain technology has an exciting opportunity to revolutionize marine resources which is otherwise considered unsustainable in nature. But the introduction of the blockchain application in local communities, especially those without basic infrastructure, faces challenges due to the lack of reliable electricity— this can be seen more significantly through private Blockchains. To address this, we propose operating a public Blockchain where mining can be performed globally via gas payments. Additionally, we plan to utilize low-power Raspberry Pi technology (OS: Raspbian) for data collection and data sharing using smart contracts.

By continuing to explore advanced technologies such as Hyperledger Aries and YOLOv8, we can improve the accuracy, safety, and efficiency of the platform for the benefit of fishermen, the Marine Office, and the marine ecosystem as a whole. In cases where fishers disagree with the use of technology or cannot afford it, we plan to implement educational programs and seek funding opportunities to support their adoption and ensure they are not left behind.

Recapitulation

The aim of this work is to produce a proof of concept on the one hand, and to work on data traceability on the other. To this end, we have tested our platform using the Ethereum blockchain. This project demonstrates the potential of technology to improve the management of marine resources, contribute to environmental sustainability, and strengthen collaboration between different industry players. With future developments and ongoing collaborations, SmartFish can play a critical role in preserving marine ecosystems and promoting sustainable fishing for generations to come. With the feasibility demonstrated in the figure, we can now explore the use of private blockchains such as Quorum or Hyperledger.

Future work

To enhance the security and privacy of fishermen’s data, the utilization of Hyperledger Aries, a decentralized identity management infrastructure, could be considered. The integration of YOLOv8, a deep learning model for object detection, could enable the system to automatically identify fish captured in images and videos, thereby reducing the workload of fishermen and improving the efficiency of data collection.

Blockchain provides an innovative approach to the sustainable management of marine resources. However, the implementation of Blockchain technology faces challenges in communities lacking basic infrastructure such as reliable access to electricity, especially in the case of private Blockchains. To address this, we propose operating a public Blockchain where mining can be performed globally via gas payments. Additionally, we plan to utilize low-power Raspberry Pi technology (OS: Raspbian) for data collection and data sharing using smart contracts.

By continuing to explore advanced technologies such as Hyperledger Aries and YOLOv8, we can improve the accuracy, safety, and efficiency of the platform for the benefit of fishermen, the Marine Office, and the marine ecosystem as a whole. In cases where fishers disagree with the use of technology or cannot afford it, we plan to implement educational programs and seek funding opportunities to support their adoption and ensure they are not left behind.