Data analytics: a game changer for maritime surveillance
The growing development of the maritime surveillance market is supposed to reach 10% CAGR from 2019 to 2026. Maritime traffic, both legal and illegal, is generating threats upon the environment, safety of men at sea and global security. While the problem of maritime authorities has been not having enough information to identify menaces, now they are flooded with so much information coming from sensors that human eyes and brain cannot analyse it all. In this context, how can AI help authorities to better anticipate and detect real threats out of regular traffic? Alcimed returns to the current maritime surveillance challenges and sketches out how the AI industry can help overcome them.
Maritime surveillance: a global growing concern
With the development of global trade, maritime traffic, responsible for most of the exchanges, is taking a crucial place in our global security system. The amount of shipped goods has increased by 4000 Million tons within a decade. Threats at sea can range from classic accidents to illegal exploitation of natural resources, smuggling, piracy, weapons traffic or human traffic and illegal migration.
Increasing maritime situational awareness, from coast to deep waters, has become a major concern for nations. The multiplication of land-based sensors, to monitor and track coastal traffic, has increased the amount of information available to maritime surveillance teams. Additionally, satellite-based asset of surveillance has greatly improved the capability of national and supra-national organization to “see” beyond radar line of sight and allowed for the surveillance of large Exclusive Economic Zones (EEZ).
From not enough to too much information: the problem of analytics in maritime surveillance
This is where data management and AI come into play and are a powerful branch of innovation able to help a maritime surveillance agent to detect threats and decide quickly which action to take.
While authorities’ main problem was being “blind” to threats or illicit activities due to the immensity of the ocean and the length of coastal line to watch for, their current problem has become having too much information.
In this context, how can one distinguish between a regular situation and a rogue one? How can an operator determine if the behaviour of a vessel is normal or not, if a trajectory is safe or potentially dangerous? The flood of information due to the increase in traffic and more technologies has blurred the capabilities of human operators to detect real threats. It is like searching for a needle in a haystack.
At the same time, maritime authorities express the need to increase their ability for early detection in order to anticipate threats rather than react to their happening. Preventing an accident rather than searching for men overboard based on vessel trajectories, early intercepting drug conveys by analysing boat speed and size pattern before actually seeing them are among the many examples of this necessary approach switch. This need becomes more important in the areas such as Europe that are suffering from downsizing, therefore loosing human capabilities to interpret information.
This is where data management and AI come into play and are a powerful branch of innovation able to help a surveillance agent to detect threats and decide quickly which action to take. Indeed, the ability to display information coming from a web of different sensors, coupled with analytics and predictive capabilities is of great help. Fortunately, it is exactly what AI does, it helps disentangle the signal from the noise based on the analysis of past patterns and situations. It can indicate to agents the most probable hazardous or rogue situations out of a tremendous number of events, that would be otherwise hidden to the human eye.
However, merging different types of data, storing them, displaying them in an integrated command and control software suite, analysing them, is a challenge given the diversity and quantity of signals and situations to account for.
Solutions are emerging in maritime surveillance but no dominant business model arises
Major historical defence players and their “maritime surveillance as a service” model
Major national defence and information technology providers, such as Thales, Lockheed or Airbus, are on the brink of industrializing solutions integrating all aspects of the problem. They provide integrated solutions, from providing sensors, airborne and shipyard platforms, to means of data analysis and AI. The business models are still, however, unclear. From selling the systems to the maritime authorities to selling modular surveillance capabilities “as a service”, all options are on the table.
By structuring their offer around the “surveillance as a service” model, major historical defence players are allowing countries with small security budgets to benefit from surveillance without having to buy expensive planes, boat or radars. However, this business model has limits and there are barriers to adoption. Maritime surveillance is often a military prerogative and armies traditionally prefer buying capital than purchasing services. Hence this new trend in business model is for the moment limited to developing countries.
Pure AI players and their data-focused solutions
Integrating the AI algorithms to user’s current command and control rooms or providing AI on user’s data as a service are some of the options proposed by companies.
Smaller companies that do not have the ability to manufacture platforms and propose a continuum of surveillance capabilities focus on bringing AI to maritime authorities. These companies are exclusively focused on data fusion and analysis. Integrating the AI algorithms to user’s current command and control rooms or providing AI on user’s data as a service are some of the options proposed by companies. Windward, ISI or the Copernicus incubator start-up SatShipAI are mid to small size companies that use such positioning.
Many models are therefore competing to address the market, from capital acquisition of solutions to analysis as a service, but none seems to have emerged as the most relevant for users. It would be necessary for a newcomer on the market, to follow the development, in the short run, of these go to market strategies.
Maritime surveillance is key for country security as a whole. The increased amount of input data coming from both the rise in traffic and sensors, has made human detection of threats highly difficult without the help of AI. The market attention has therefore drifted from providing sensors to providing fusion and analysis of these sensors’ data. From major defence companies proposing a continuum of surveillance capabilities to data-focused companies focusing on bringing AI to users, many models are competing, and none is dominating yet. Should we expect a vertical integration of the market and pure players be purchased by major defence providers? Will the pure players have the size to compete with the majors in the long run? All routes remain open in this emerging market and choosing the right business model will be key to address it.
Alcimed is passionate about exploring new go to market strategies and mare incognita. The rise of AI in maritime surveillance is a real challenge all along the value chain, that we will keep watch over.
About the author
Alexandre, Business Unit Director in Alcimed’s Aeronautics Space Defence team in France
Do you have an exploration project?
Our explorers are ready to discuss it with you