The study presented an analysis of the on-deck sorting selection of pair trawlers for three species: John dory, whitefin trevally, and yellowback seabream. The previous paper revealed that small individuals of these species were caught with trawl codend of the legal minimum mesh size due to their large body height relative to body length. The sorting selection curve for landing was estimated using a logistic curve equation through the SELECT analysis, with a comparison of body length composition between landed fish measured at the landing site by the pair trawlers and fish collected at trawl surveys at the same fishing ground. For John dory and yellowback seabream, the landings were clearly sorted by body size, with a pronounced narrowing of the landing selection length ranges. In particular, the 50% landing selection length of 12.3 cm for yellowback seabream was relatively close to the fish length of 99% retention by the codend of the legal minimum mesh size. This suggests that the majority of fish captured by pair trawl was landed. In contrast, the fish lengths of 50% landing selection for John dory and whitefin trevally were larger than the fish lengths of 50% retention by the codend, indicating that a considerable number of small individuals were discarded. Furthermore, the potential for overlooking growth overfishing when conducting yield per recruit analysis with data on fish length obtained exclusively from landings at fish markets was also discussed.

Research workshops were held on 24 May 2025 during the annual meeting of the Japanese Society of Fisheries Engineering. The purpose of holding research workshops was to provide opportunities outside the main lecture sessions for participants to spend more time-sharing information and exchanging opinions within each field of fisheries engineering. In addition, as the conference was being held in a regional area, we hoped to create opportunities that will encourage as many members as possible to take part. And this set of discipline specific research workshops provided a valuable opportunity to share the latest insights from each field and to explore possible directions for future research.

Designing fishing port facilities such as breakwaters involves verifying whether the structure remains stable and safe for use against external forces like typhoons, high waves, and earthquakes. While existing design methods are updated annually based on new insights into phenomena, advancements in analysis techniques, and other factors, new design approaches are also frequently developed in response to lessons learned from unique or unprecedented natural disasters. This paper reviews part of the design methodology developed in response to the tsunami damage caused by the Great East Japan Earthquake of 2011, the most impactful natural disaster in recent years. It also introduces methodologies and concepts for adapting to already apparent societal and environmental changes. Furthermore, it suggests perspectives on the direction of technological development to prepare for anticipated future changes. Japan’s geographical location, topography, geology, and meteorological conditions make it prone to disasters such as typhoons, earthquakes, and tsunamis. Existing design methodologies have been developed and refined based on lessons learned from past natural disasters. While concerns persist about new natural disasters arising from climate change impacts, it remains crucial that once a disaster occurs, its causes are identified and design methodologies are updated to prevent recurrence.

The formula for calculating hydrodynamic forces on submerged artificial reefs was overhauled in the 2023 revision of the Design Reference Guide for Japan Fisheries Infrastructures. The new formula has been verified to estimate maximum forces with reasonable accuracy despite its simplicity, even under complex conditions such as asymmetric velocity fields generated by breaking waves, and combined wave-current fields where waves and a current intersect at arbitrary angles. To apply the formula, it is only necessary to determine two parameters: the drag coefficient in a steady current (C_DS) for the target reef, and the maximum force coefficient (C_Fmax), which is expressed as a function of the Keulegan-Carpenter (KC) number. C_Fmax decreases monotonically with increasing KC number, asymptotically approaching C_DS. By utilizing this characteristic, a reliable empirical formula for C_Fmax can be easily established using C_DS. Submerged artificial reefs come in many different shapes. However, a significant number of these reefs are constructed from members with simple cross-sectional shapes, such as square prisms and L-shaped angles, with wide spacing between them. This suggests the possibility of the ‘Component Force Method’ (CFM) or the method of determining the total hydrodynamic force on the entire reef by summing the forces acting on its individual components. While the CFM has been verified in a simple structure reef, it faces challenges when members are inclined to the flow or when the spacing is not sufficiently wide. As a solution to the latter issue, the cross-flow principle has been tested, and its applicability has been demonstrated.

In recent years, Japan’s fisheries production has declined due to resource depletion and changes in the marine environment, bringing the industry to a major turning point. The Fisheries Agency has analyzed the causes of poor catches and advanced measures to address climate change, identifying the diversification and conversion of fishing methods and target species as key directions. JAMARC has conducted a wide range of pilot studies, including extending fishing seasons and grounds, utilizing underused and unexploited resources, enabling squid jigging by Pacific saury fishing vessels, and improving fishing gear to cope with complicated ocean currents. These efforts contribute to increasing vessel utilization rates and securing new income sources, offering practical strategies to reduce dependence on single-species fisheries. Under accelerating climate change, traditional experiential knowledge alone is insufficient; innovation in fishing gear, commercialization of underutilized resources, the adoption of AI and ICT technologies, and collaboration among relevant institutions are essential. Supportive governmental frameworks are also required, as regional ingenuity and proactive adaptation are crucial for sustainable fisheries management.

Recent advances in artificial intelligence (AI) and image processing have enabled automated analysis of underwater videos in a wide range of fisheries and marine science applications. AI-based approaches have been actively applied in laboratory experiments, ecosystem monitoring, and aquaculture. In contrast, their application to fishing gear and fishing method research remains relatively limited. This article reviews recent studies on AI-based video analysis in laboratory tanks, ecosystem monitoring, aquaculture, and fishing gear and fishing method research, and discusses the factors that make the application of AI particularly challenging in fishing gear and fishing method research. In addition, image processing techniques to improve the efficiency of underwater video analysis in set nets and bottom trawl fisheries are introduced. Edge detection is used to extract candidate time periods with fish presence in set net videos, while a frame differencing method is employed to detect fish and marine debris in bottom trawl videos. These approaches effectively reduce annotation workload and facilitate efficient training data preparation. The integration of image processing and AI is expected to accelerate fishing gear research and contribute to the development of sustainable fisheries.

From FY2022 to FY2023, preliminary designs for zero-emission fishing vessels were conducted for five vessel types across three offshore fisheries: two types of purse seiner, two types of offshore trawler, and one type of pacific saury fishing vessel. These designs serve as initial models to support future decarbonization efforts, though they are not directly linked to shipbuilding. From FY2023 to FY2025, a fuel cell-powered feeding boat is being developed, constructed, and tested at sea. The design is based on simulations of energy consumption during feeding operations, ensuring current work efficiency while meeting equipment requirements. Safety is ensured in accordance with relevant guidelines. To implement fuel cell fishing vessels in the fisheries sector, it is essential to resolve technical challenges, establish economical fuel supply systems, and develop crew qualification frameworks and safety standards. These initiatives represent pioneering efforts, and their outcomes are expected to contribute to the practical use of hydrogen energy in fishing vessels.

Initiatives toward a carbon neutral society have been promoted by international societies, especially by the IPCC (Intergovernmental Panel on Climate Change). As one renewable energy resource with high potential, offshore wind energy has been given a special focus over the past decade. In Japan, there is an exceptional expectation given to offshore wind energy as a core resource, taking advantage of the vast oceanic area around the country and its strong wind potential. It is against this backdrop that Toyo Construction, a Japanese marine contractor, has been conducting research and development activities, investing in new assets and introducing technology such as an offshore construction vessel and related equipment which are associated with offshore wind power plant construction. In this paper, the current situation and the prospect of offshore wind energy are overviewed in both global and domestic perspectives. In the latter half of this paper, some examples of Toyo Construction’s approaches are summarized.

Bluefin tuna are naturally fast-swimming fish. However, the physiological stress caused by rearing them in the confined spaces of aquaculture net pens, combined with overcrowding, can lead to collisions with the nets, resulting in injury or death. This stress may also cause abnormal behaviors, such as attempts to escape from the pens. Understanding the swimming behavior of bluefin tuna in net pens is essential for preventing accidents caused by abnormal behaviors. Over an eight-year period from 2017 to 2024, the authors employed multi-transmit/receive sonar and pingers to count the number of fish in net pens and to measure the swimming behavior of bluefin tuna in 116 pens operated by 15 aquaculture companies across Japan. This study focused on the swimming speed of bluefin tuna, using 3D imaging to analyze variations in swimming speed related to seasonal changes, net pen size, and sequential changes from daytime to twilight and nighttime. The goal of this research is to provide foundational information for creating suitable aquaculture environments that consider the comfort of farmed fish by clarifying the behavioral characteristics of bluefin tuna in net pens.

The 2025 JSFE Spring Symposium, entitled “Future Perspectives on Automation Technologies in Fisheries” was held on 25 May 2025. In recent years, the demand for automation and smart technologies in the fisheries industry has been steadily increasing. To address labor shortages and the growing complexity of operations, many regions are adopting AI- and IoT-based systems to improve production efficiency and enhance safety. Against this background, this symposium presented current trends in automation and smart technologies within not only in fisheries, but also in the agriculture, and livestock industries, while also highlighting the rapidly evolving developments in generative AI and related technologies.

The Japanese fishery industry faces critical challenges, including labor shortages and inefficiencies due to analog practices. This paper reports on the construction of a digital data linkage system for offshore trawling fisheries in Shimonoseki, characterized by initiating digitalization directly at the production site. A significant achievement of this system is the dramatic condensation of work time through automation. By introducing the “Fisheries Support Application,” which incorporates incentives like real-time revenue forecasts to ensure accurate data entry, labor-intensive tasks such as aggregating catch data previously requiring approximately 17 hours per voyage were fully automated. Additionally, statutory reporting time was reduced from one hour to three minutes, allowing fishermen to focus on core operations. Furthermore, seamless data linkage to the “Market Application” establishes “information precedence,” where catch details reach the market before the vessel docks. This allows wholesalers to formulate sales strategies in advance and automates sales slip creation. Consequently, the system not only resolves information asymmetry but also transforms the time saved into resources for strategic decision-making and high-value-added operations, demonstrating a model for optimizing the entire supply chain and ensuring sustainable fishery management.

This paper investigated the integration of Large Language Models (LLMs) with Multi-Agent Deep Reinforcement Learning (MADRL) to enhance autonomous coordination in marine environments. While Generative AI is evolving into autonomous multi-agent systems, coordinating these agents remains a challenge. The study proposed a “Marine Debris Collection Game” where vessel agents utilize MADRL for tactical navigation and LLMs for strategic communication regarding departure decisions. Simulation results demonstrated that while MADRL improved collection efficiency compared to random movements, the addition of LLM-based dialogue reduced fuel consumption by optimizing fleet deployment without compromising total collection volume. Although a disconnect remains between strategic LLM decisions and tactical control, this model suggests promising applications for automated fisheries and marine infrastructure maintenance, demonstrating how language-based negotiation can facilitate resource-efficient cooperation in multi-robot systems.