Measuring impact is one of the challenges in sustainability work.
Take seaweed farming. Knowing how much seaweed is growing below the surface helps a farmer plan a harvest and decide where to invest. Through a project supported by Canada’s Ocean Supercluster, Coastal Carbon and HoldFast are developing an AI-enabled sensor to monitor seaweed growth remotely. It is a practical example of how AI can help address a measurement challenge in the ocean.
When you can measure something properly, you can build a business case for it. Evidence gives people a reason to invest.
A 2023 BCG report commissioned by and co-authored with Google estimated that scaling existing AI applications could help mitigate 5–10% of global greenhouse gas emissions by 2030.
That is an estimate of potential impact. Realizing it depends on how widely and effectively those applications are used. What interests me is how better information can help us make better decisions, and how we measure the difference those decisions make.
We have so much of the ocean where data is not collected yet, and then we have so much data that we haven’t figured out how to analyze.
Imagine a fully instrumented ocean, where we understand what’s happening in real time. Environmental events, biodiversity movement, illegal fishing, ship emissions, tsunamis. Pick your ocean-related problem, and many get better with better information.
What we're really trying to do is accelerate the path from data to information to insight. We're collecting more data than ever, which is giving us more information than ever. Turning that into insight, and then into decisions, is where we're still early in the journey.
Food production offers a practical example. In a study of European seabass, researchers used AI and computer vision to monitor swimming behaviour under different feeding conditions. That work can help inform decisions about when fish have had enough to eat. It points to an opportunity to match feeding more closely to what the fish need.
In fisheries, NOAA scientists are using AI to process video data more quickly to evaluate devices that help salmon escape from nets intended to catch pollock. That gives researchers information they can use to assess and improve ways of reducing unintended catches.
Graphite Innovation and Technologies, a Nova Scotia company whose coating project received funding from Canada’s Ocean Supercluster, illustrates why measurement matters. Its graphene-based coatings are designed to reduce friction and improve vessel efficiency. Assessing that performance in service requires attention to the vessel’s operating conditions and how its hull is maintained. For a shipping company considering an investment, the useful question is what changes in its own operations and how reliably that change can be measured.
Better weather prediction helps too. Even a small adjustment to a ship’s route can reduce fuel consumption. In ports, AI can improve how vessels are berthed and reduce unnecessary container movements around the yard.
Offshore energy presents another opportunity because AI can help manage wave and wind installations more efficiently. Biodiversity protection begins with understanding what is actually below the surface.
It’s pervasive. AI is already being applied to food, transport, energy and biodiversity. It can improve the decisions we make and help us measure the result. That makes the next investment easier to justify.
Innovators understand the technology. Businesses understand the problems. The magic, to me, is bringing the two together around a business case.
That collaboration matters when we move from developing a tool to testing it in practice. An AI model needs relevant data, and the people who understand the operation help determine whether its output is useful. Leaders need to know what has changed and whether it is enough to justify the next investment.
Many companies bring in a nimble technology firm instead of building that capability internally. The solution comes from their work together. That is also the foundation of our model at Canada’s Ocean Supercluster.
Start with a specific operational problem, then find the right people to work on it and be clear about the data you are prepared to share.
Data collection is advancing quickly. Sensors, drones and autonomous platforms can reach parts of the ocean we’ve never accessed before. Low Earth orbit satellites have made access cheaper and more frequent. Access to data from places that have historically been difficult to monitor, including much of the Arctic, is improving. The volume of data will be huge.
That raises harder questions about who controls the data and how it is shared. Sovereignty is part of that, especially in the North. Leaders need to be clear about which data can be shared, with whom, and for what purpose. Those decisions are part of making collaboration possible.
Ultimately, more than 70% of the world’s seafloor remains unmapped to modern standards. I find that exciting. What we find could expand our knowledge of ocean species, point to new health solutions and improve coastal resilience.
For leaders, the starting point is articulating the problem clearly enough to know which question to ask. We are about to have access to more ocean data than ever. AI can help us make sense of it and find answers that were out of reach before.
Kendra MacDonald is CEO of Canada's Ocean Supercluster, a federal innovation cluster focused on growing Canada's ocean economy. A recipient of the King Charles Award for her contributions to Canada, she has been named one of Atlantic Business Magazine's Top 50 CEOs for five consecutive years and was inducted into the Top 50 CEOs Hall of Fame. She is a member of the International Women's Forum and serves on the Board of the Canadian Chamber of Commerce.
