At restaurants and dinner tables around the world, seafood is often the entrée of choice. Fish, crustacean, and mollusk consumption account for about 17 percent of the world’s total animal protein intake, with much of this coming from the ocean. Fish and shellfish are especially important in low-income areas where total protein intake is low and diets are less diversified.
Fishing companies—businesses that catch fish or other seafood in the wild—will play a major role in sustaining food security and supporting fishing communities. But in their quest to capture enough fish to satisfy soaring demand, they are exerting unprecedented pressure on marine and freshwater ecosystems. It now takes five times the effort (in kilowatt-hours) to catch the same amount of fish as it did in 1950, because the targeted species are now in scarce supply.1 This shortage not only jeopardizes commercial prospects for fishing companies but also greatly threatens the ability of endangered ocean species to reproduce and maintain their numbers.
Balancing fishery interests with environmental concerns is not easy, but advanced analytics (AA)—the use of sophisticated methods to collect, process, and interpret big data—might represent an untapped solution to this problem. While fishing companies, regulators, and environmentalists now apply these tools, their use is typically limited to small-scale pilots. But we may have reached the point where advanced analytics will take off within the fishing sector. In addition to the development of new technologies that support analytics in this field, both policy makers and fishing-company leaders have an increased sense of urgency because of dwindling fish stocks. Further, people entering the fishing industry or participating in regulatory development are more tech savvy than their predecessors, giving them a greater understanding of advanced analytics and other digital tools. Even fishermen from emerging markets can access information on these technologies—and their benefits—through a simple smartphone search.
The growth of advanced analytics could promote the development of precision fishing—the use of advanced tools and technologies to optimize fishing operations and management. If large-scale fishing companies around the world move to this model, they could decrease their annual operating costs by about $11 billion, and customers would benefit from lower prices for fish and seafood. Precision-fishing techniques can also contribute to improved management of ocean resources, which could increase industry profits by as much as $53 billion by 2050 while simultaneously raising the total fish biomass to at least twice the current level.2
This article attempts to paint a picture of the current situation in the fishing industry, focusing on the challenges that are making it more urgent to adapt advanced analytics and associated tools. It also discusses several of the most popular use cases that have emerged for advanced analytics, as well as others that show great potential. Finally, the article provides a practical guide to next steps for all industry stakeholders.
The appetite for tuna, salmon, shrimp, and other ocean creatures is nothing new. Demand has increased an average of 3.2 percent annually between 1961 and 2016—more than twice the 1.6 percent rate of population growth over the same period and higher than the 2.8 percent rise in consumption of terrestrial mammals.3 Overall, the world’s fish consumption is predicted to increase by 20 percent from 2016 to 2030, driven by global population growth, the expansion of the middle class, and greater urbanization (giving more people more access to seafood, as well as the electricity and refrigeration needed to store it). Consumers also increasingly prefer healthy food choices, and many view fish as a good alternative to red meat.
As boats across the world search for a good haul, wild-fish capture has been slowly declining. Since the mid-1990s, the amount of wild fish processed has fallen by about 0.6 percent on an annual basis, while the amount coming from aquaculture rose by 5.7 percent (Exhibit 1). (Aquaculture production comprises entities that breed, rear, and harvest all types of fish as well as other organisms that live in water.) The value of fish coming from aquaculture now tops $250 billion annually, compared with about $170 billion for wild catches.
To cope with the decreased catch in their traditional fishing grounds, commercial fishing companies have considerably expanded their footprint on the oceans. In addition to targeting new species, they have increased their fishing efforts in tropical zones and extended their operations from coastal regions to the high seas, raising the total area fished from 60 percent to 90 percent of the world’s oceans.4
Overall, the world’s fish consumption is predicted to increase by 20 percent from 2016 to 2030, driven by global population growth, the expansion of the middle class, and greater urbanization.
Thanks to technological improvements, fishing companies have also penetrated further depths to target deepwater animals such as grenadiers and blue lings. Fishing these species is rarely sustainable because many have slow reproduction rates, which limits spawning and population growth. In the past, targeting such fish has often resulted in ecological disasters. In the 1980s, for instance, the deep-sea orange roughy almost suffered extinction through overfishing until researchers discovered that it was slow growing and exceptionally late to mature.
As fishing companies expand their reach, they are putting extreme pressure on the ocean environment. About half the world’s fish stocks are now classified as collapsed, rebuilding, or overexploited, and wild-catch rates are falling in most regions (Exhibit 2). This phenomenon is particularly apparent with large fish at the top of the food chain, including sharks, tuna, and billfish.5 The loss of these apex predators has cascading effects that disrupt the equilibrium of ocean ecosystems.6 Take the decline of some shark populations, which has been known to trigger sudden and undesirable population changes in species living in the same habitat. The number of shellfish or herbivores might collapse, for instance, or a large algae bloom could develop.
Other perils also loom. By 2025, oceans could contain 250 million metric tons of plastic—one per every three tons of fish—unless companies and other stakeholders institute some mitigation measures.7 The accumulation of plastic debris may reduce the fish-survival rate, lowering stocks. Climate change, and its accompanying acidification, warming, and deoxygenation processes, is already affecting the oceans and will have profound implications for marine ecosystems, including reduced biodiversity and shifts in habitat. According to some scenarios, these shifts could decrease fishing revenues by 35 percent by 2050.8
Recognizing the growing threat to fish stocks, some countries and regions have acted to improve resource management, with mixed results.9 For instance, the United States has increased the proportion of stocks fished at biologically sustainable levels from 53 percent to 74 percent from 2005 through 2016, an increase that may be partly attributed to the Magnuson-Stevens Fishery Conservation and Management Act.10 Similarly, around 69 percent of stocks managed by the Australian Fisheries Management Authority were sustainably fished in 2015. But these regional gains are negated by overfishing in other markets, illegal fishing, and excessive waste.
Since regulations alone cannot eliminate overfishing, fisheries need other solutions to stay on a sustainable trajectory while minimizing their environmental impact. For most issues, including catch reporting, trade-information sharing, subsidies, tariff policies, and regulation enforcement, greater national and international collaboration will help. But fisheries and the public could also benefit from the increased use of advanced analytics (Exhibit 3). These algorithms have become popular across industries over the past few years as technological improvements have increased data availability, facilitated the deployment of information, and expanded data-ingestion capabilities.
Many industry stakeholders have already incorporated advanced analytics into all components of the value chain. Here’s a look at some of the most important recent developments relevant to fisheries.
Data acquisition through sensing platforms
Sensors for collecting data have become more common, compact, and less expensive over the past few years. At the same time, the variety of platforms on which these devices can be deployed has considerably expanded, allowing them to capture data more rapidly and over greater distances. Sensing platforms that are particularly important within the fishing industry include the following:
- Satellite. Optical and radar sensors on satellites can offer a holistic view of the environment at unprecedented spatial and temporal resolution, making them particularly valuable for monitoring purposes. Optical sensors measure the light reflected by the earth’s surface across a wide range of the electromagnetic spectrum. Important oceanic parameters can be derived from such data, including sea temperature and turbidity. Radar sensors emit microwave radiation and measure the portion that is scattered back to the instrument. They can provide data about ocean topography, winds, sea ice, and the movement of vessels. Unlike optical sensors, radar systems can collect information even during poor weather and lighting conditions, including times when the sky is dark or cloudy.
- Drones. Equipped with cameras or other sensing devices, drones are increasingly used to explore the ocean. Some are even capable of navigating underwater. Compared with oceanographic vessels, drones are cheaper and more flexible. When sent in groups, they can also provide a more exhaustive sampling of the environment.11 Although drones cover a smaller area than satellites, they can provide more detailed images, allowing them to detect smaller objects or phenomena.
- Onboard or underwater devices. Data related to fishing operations and catch are typically recorded by fishermen or observers. Common parameters include those related to vessel location, gear types, and catch, including species, volume, biophysical characteristics, and discards. Onboard sensors can automate and facilitate this laborious process while simultaneously generating more exhaustive and reliable data. The data are then integrated into platforms known as electronic monitoring systems (EMSs). Several fishing-management authorities also require large fishing vessels to be equipped with vessel-monitoring systems (VMS), a technology that the European Union established in the early 2000s to support the monitoring, control, and surveillance of fishing vessels in its waters. VMS can collect information on a vessel’s position, speed, and heading. Vessel operators can also send valuable information to authorities through their VMS, such as estimated catch and the start and end times for their fishing operations. Another onboard utility, the automatic identification system (AIS), was designed to complement radar systems and decrease the likelihood of marine collisions. Like VMS, it can be used to track the activity of fishing vessels. Other sensors, such as cameras and fuel-monitoring systems, can also be placed on board or next to underwater nets for real-time tracking.
Public organizations such as the National Oceanic and Atmospheric Administration and the Copernicus Marine Environment Monitoring Service have increased the effective usage of data obtained from satellite sensors by freely publishing them. Many start-ups and other companies also offer various products related to sensing platforms, including output from satellite sensors and data-collection systems designed for commercial fisheries.
Improved data-transmission technologies
The growth of the Internet of Things (IoT), land- and satellite-based mobile networks, and smartphones makes it much easier for fisheries to transmit data from vessels for analysis. For instance, vessels can use IoT to monitor and transmit data on fuel consumption in real time. The resulting data are then sent ashore through wireless mobile networks, including 3G and 4G, when close to shore. At further distances, vessels rely on satellite networks for transmission.
The growth of the Internet of Things (IoT), land- and satellite-based mobile networks, and smartphones makes it much easier for fisheries to transmit data from vessels for analysis.
More insightful data analysis
Computational power has increased substantially, making it easier to process and analyze information using sophisticated algorithms. Across industries, some of the most important advances relate to the rise of artificial intelligence and machine learning, which can identify hidden relationships in large amounts of data. In particular, image-recognition and object-detection tools, powered by deep learning, have made a significant leap forward during the past decade. For instance, onboard cameras, assisted by image-recognition software, can provide fishermen with important information on the content of their catch in real time, including species, volume, and fish size.
Fishing-industry stakeholders are already transforming their operational and business processes by incorporating AA into all parts of the value chain, including fishery management, detection and capture, processing, reporting, and surveillance and control (Exhibit 4). They typically use multiple AA tools and sensors in combination, and a few even apply them from end to end within the value chain (see sidebar, “How are fisheries exploring new technology? An interview with Matts Johansen, CEO of Aker BioMarine Antarctic”). We have found that in some of the most important use cases involving AA and fishing, the following actions have been taken.
Monitoring illegal, unreported, and unregulated fishing
Authorities leverage AA to combat illegal, unregulated, and unreported fishing using geolocation data from AIS and VMS. AA can predict whether fishing vessels are actively engaged in fishing by looking at their AIS speed and course profile. For example, a vessel that slows down to one to three knots and frequently changes direction would likely be fishing. If geolocation data are not available, AA can also determine the position of vessels through image-recognition algorithms and satellite imagery (both radar and optical) that allow authorities to monitor the fishing fleet directly from space and detect any suspicious activity under their purview, such as fishing in restricted zones or the offloading of fish cargo from one vessel to a refrigerated transport vessel—a practice that is sometimes used to conceal a catch from authorities.
Some industry organizations also use sensor data to monitor fishing activity, with the goal of increasing sustainability, such as Global Fishing Watch, a not-for-profit organization that aims to increase transparency by offering free data about the activity of the global fishing fleet based on AIS, VMS, and satellite imagery.12
Improving the detection of fish
Most fisheries have scarce data about their target catch. They might assess stock yearly, rather than making more frequent observations, and their analyses focus on information about landed catches and data recorded by observers. Tools that incorporate advanced analytics can provide a more dynamic, reliable, and nuanced view of the fluctuating ocean environment.
Consider patterns related to fish aggregation and migration, which change in response to temperature, wave height, the presence of sea ice, and other ocean conditions. Fisheries can monitor these changes through satellite imagery obtained from sensors. Complemented with information from other sources, such as the location of fishing vessels and catch data, advanced analytics can help determine the distribution and migratory patterns of a target species over time and space with greater accuracy and frequency.
Some researchers have already applied advanced analytics to get better information on the distribution of fish. One team developed high-resolution predictive models by combining various ocean data, including sea-surface temperature, wind speed, and chlorophyll levels associated with plankton, with information obtained from fisheries and tagging sensors. The models provide daily recommendations about where to fish and how to avoid bycatch, increasing efficiency.13 With a more detailed and dynamic vision of fish stocks, fishing companies can decrease the amount of time, effort, and fuel required for each catch. Likewise, authorities can use the data to improve resource management.
Reporting to authorities and central managers
As noted earlier, fishermen and independent observers typically monitor and report fishing activities themselves. The results are then sent to relevant authorities or central managers within their company. EMS can automate and facilitate this time-consuming process to generate more exhaustive and reliable data based on sensor input. These systems typically consist of cameras connected to a GPS receiver and other vessel-tracking devices, such as engine-monitoring sensors that send data on fuel consumption in near real time. As fishing companies evolve toward a more data-rich environment, advanced analytics will become more and more relevant. Eventually, fishing companies will be able to combine data in ways that deliver new insights about key operational-performance drivers, such as fuel consumption and fish-catching rates.
The supply chain in the seafood industry is complex, opaque, and lacking in international harmonization because the stakeholders involved often closely guard their information. 14 The lack of clarity makes it easier for vessels to skirt regulations and fish illegally. It also frustrates consumers, who are increasingly asking for more information about the source and freshness of the food on their plates.
The supply chain in the seafood industry is complex, opaque, and lacking in international harmonization because the stakeholders involved often closely guard their information.
To improve transparency, some researchers are investigating distributed-ledger technologies that track and store information on transactions, including data on the movement of goods along the supply chain, in a secure, distributed database. Although distributed-ledger technologies are not classified as advanced-analytics tools, they are an important enabler. The information in a distributed-ledger technology database, including insights from advanced analytics, is available to all approved users in real time.
Researchers are also investigating other technologies for tracking seafood, such as radiofrequency-identification tags and quick response codes, both of which transmit product information when scanned. With tagging, fishing companies may find it easier to receive permission to place labels on their products certifying that they are approved by the Marine Stewardship Council and other organizations that guarantee a product has been sustainably sourced, monitored along the supply chain, and correctly labeled. Consumers may increasingly look for such labels, giving an advantage to those that fish responsibly.
Although commercial fishing companies are exploring advanced analytics through pilots and other activities, their decisions about where, when, and how to fish are still largely based on intuition and experience. Similarly, most regulators are not taking full advantage of advanced analytics. They have collected and analyzed some data, but their information is often incomplete and prone to inaccuracies, especially in emerging markets.
With all industry stakeholders concerned about fishing stocks, it is now time to take a more aggressive approach to advanced analytics. As noted earlier, recent technological advances will facilitate this push, since costs for data storage and processing are decreasing each year. Their greater affordability means that most fishing companies and other stakeholders can now afford to implement more advanced-analytics tools in the near future. Likewise, talent recruitment will become less difficult for fisheries since the supply of data scientists, engineers, and technicians is growing. Fisheries will still face more challenges in acquiring talent than well-known tech companies or other industries that have traditionally promoted advanced analytics, but the recruitment pool will be larger.
To guide their advanced-analytics journey, fishing companies must create a road map focusing on challenges they hope to address, such as those related to fishing efficiency, capture volatility, and fleet monitoring. To identify quick wins, companies should first assess their data stores to see what information is readily available. Most will find that they already have much relevant information on hand, including vessel-specific data on daily catch (both volume and species), GPS position, and fuel consumption.
Although commercial fishing companies are exploring advanced analytics through pilots and other activities, their decisions about where, when, and how to fish are still largely based on intuition and experience.
Simple yet powerful use cases could be built around such data. Rather than using this information for purely descriptive purposes—for instance, noting the average catch for each vessel during past months—fishing companies could adopt a forward-looking analytical approach. One analysis might involve using geospatial modeling to map fishing activity and catch rate over the course of the season, allowing fisheries to track the fleet more closely and gain a better understanding of performance drivers. Increased fishing efficiency would also reduce fuel consumption and running costs. In addition to such simple analyses, fishing companies could use geospatial modeling to predict the location of targeted fish according to various environmental conditions. Such tools could inform not only fishing operations but also downstream commercial activities, including seafood pricing and labeling.
Fishing companies will also find many other use cases for advanced analytics. For example, they could generate even greater fuel savings by examining data from IoT sensors that provide information on vessel behavior, including fuel consumption and navigation conditions. Their analyses could help them generate real-time recommendations about the most energy-efficient routes and maneuvers. Similarly, fishing companies could examine data from onboard sensors to determine if any equipment is experiencing the sorts of problems that typically occur before a breakdown. With this information, they could detect potential failures ahead of time, thereby preventing costly repairs and long downtimes. In an analysis of large fishing companies worldwide, we estimated that using advanced analytics could produce more than $11 billion in savings by reducing running costs, as well as expenses for fuel, labor, and repair and maintenance (Exhibit 5).
While these potential gains are impressive, fishing companies will not achieve them by simply implementing advanced-analytics initiatives. Instead, they must undertake an end-to-end digital transformation throughout all their functions.15 Such transformations require employees to have the right skill sets, as well as appropriate tools, processes, and interfaces (for instance, dashboards where they can readily access data). In addition, organizations should provide training and support to help employees see the value of advanced analytics, especially if they appear reluctant to change their ways. Without this support, employees may view advanced analytics as an imposition—a mind-set that is likely to impede progress.
Government and fishery-management agencies
Advanced analytics can increase transparency about seafood from ocean capture to the dinner table
With fish stocks dwindling and environmental challenges mounting, governments and fishery-management agencies could consider investing in data-collection technologies and research programs that can provide a comprehensive, near real-time vision of both ocean resources and fishing activities. By leveraging the data, they can adopt new measures and regulations more quickly and also rapidly respond to external pressures such as climate change. Fishing quotas could also become more dynamic. Rather than setting a quota annually, at the beginning of the fishing season, authorities could make adjustments throughout the year based on real-time information about the amount and type of catch that vessels are collecting.
The current exchange of information between fishermen and authorities is not optimal.16 A collaborative problem-solving approach—potentially happening at the global or regional level—is needed to develop a clear road map defining data standards and mutual goals, such as those for by catch reduction. These efforts would build trust among stakeholders and benefit all.
By improving both the monitoring of fishing activities and the reporting of associated catches, advanced analytics can increase transparency about the seafood supply chain from ocean capture to the dinner table. Food companies can share this information with consumers, who have a growing interest in the quality, traceability, and sustainability of food products. In addition to their own health, they are concerned about environmental impact. If advanced analytics reveals that most of a company’s catch comes from endangered species or overfished areas, the company can shift to other options to increase sustainability (either moving its own fishing fleets or changing suppliers). Certain technologies, including distributed-ledger technologies and radio-frequency-identification tags, can help companies share their insights about catch origin more efficiently and might merit additional investment.
Modern farmers already rely on sophisticated weather forecasts, sensors, and geospatial tools to optimize their harvest and manage land more sustainably. Now it’s time for fishing companies and other stakeholders to start their own digital and analytical journey. Getting fuller nets and larger fish is one goal, but a more important objective relates to sustainability. As fish stocks drop and fishing companies expand their reach, advanced analytics may be one of the best tools for protecting endangered species and other ocean resources. While data and algorithms may seem a better fit for boardrooms than boats, some fisheries have already achieved major gains by applying them. It’s now time for more widespread adoption before the environmental consequences of overfishing accelerate.