Satellites vs. Plastic Why Tech Alone Won’t Fix It

By Darwish Thajudeen

“What if we could spot ocean plastic from space?”, this question is no longer science fiction. Satellite imagery and AI are now being used to detect floating marine litter, map hotspots, and support cleanup planning. But the more important question is this: Does being able to see the problem mean we would actually be able to solve it?

The answer is: not by technology alone.

At MI4People, our Marine Litter project is exploring how AI supported computer vision can help identify floating debris from satellite imagery and turn it into a free, publicly accessible map for researchers, cleanup groups, and environmental advocates. Which has been a meaningful step forward. Yet it also reveals a deeper truth about environmental innovation; detection is not the final solution.

What satellite imagery combined with AI does well?

Satellite monitoring gives us scale and speed, which is the most crucial factor, that marine pollution has long lacked. Instead of relying only on expensive, local surveys, we can now scan large ocean areas repeatedly and identify likely concentration zones over time. This matters because plastic pollution is dynamic due to the currents, winds, storms, and coastal geography, which all can influence where litter accumulates.

Recent research (ESA, 2024) has shown that satellite based detection of marine litter is increasingly feasible, especially when debris forms dense windrows or surface patches large enough to be visible from space. For conservation teams, that means satellites can help prioritise where to look, where to verify, and where cleanup efforts might have the highest return, within the shortest time.

The first time we detect a cluster of floating debris from space is exciting. It feels like a magic solution that can tells us exactly where to clean. But the real challenge begins once we stop looking at pixels and start looking at the real world.

Even with perfect detection, most of the plastic in our oceans does not float in big patches. A 2025 meta analysis estimates the total mass of marine microplastics already exceeds 1.5 billion tonnes. This has far exceed previous expectations and projections. Microplastics are now present in almost all marine habitats, from the surface to the deepest abyss, at an average global density of around 38,000 particles per square metre. That is a dispersed, microscopic storm that no satellite or clean up vessel can fully contain.

Moreover, satellite systems can only detect what is visible at the surface, under favourable imaging conditions, and often only when litter is sufficiently concentrated. Clouds, sun glint, wave patterns, foam, seaweed, and even ship wakes can all interfere. Also, even if a model is highly accurate, it still cannot remove plastic, redesign packaging, improve waste collection, or enforce producer responsibility. Which is the real challenge. We believe marine plastic pollution is not just a monitoring problem. It is a systems problem, that needs system solutions.Insights from the Marine Litter project

One of the most interesting insights from our work is that hotspots are not always where we expect them to be. Some accumulation zones emerge because of local currents, weather events, and coastal dynamics rather than population density alone. In fact, many communities that suffer the most from plastic deposition are not near major river outflows or densely populated coastlines, but in small bays where local currents and seasonal storms create fleeting, yet severe, accumulations. Often times these populations might not even be the producers of the pollution, but mere victims.The clean up space is full of glossy high tech contraptions – drones, robots, floating booms – that promise to “solve” plastic pollution. Yet a 2024 Royal Society report concluded that many of these technologies lack evidence on their effectiveness, scalability, and negative environmental impacts. Simple, low tech solutions like community beach clean ups are often more effective at limiting harm to humans and ecosystems than unproven high tech devices.In other words, technology isn’t the complete solution, it can only tell you how and where to look, but stopping the actual problem happens upstream, on land, in product design, and in how we manage waste.AI is a mere decision support tool that surfaces unexpected patterns and helps reveal where intervention may matter most. But actual prevention and collection requires human efforts and support of every individual on this planet. We along with our supporters are trying our best in this effort by delivering an open and transparent systems. By making the Marine Litter prototype publicly accessible, MI4People is not just building a model. It is building a shared environmental resource that can support validation, collaboration, and better targeting of cleanup action.

We have built a prototype that can scan satellite images every two to five days, identify floating plastic, and track it over time. That is a leap forward. However, we will fail if we treat AI as the victor and close our laptops. Our collaboration with Alexander Thamm and its pro bono data scientists shows what volunteer passion and professional expertise can achieve together. But we have also learnt that even the best models are useless if no one on the ground acts on them, because true success requires:

- Open data and local verification. Pixels must be checked by people who understand tides and seasons.

- Prevention first. The most effective cleanup is the one that never happens. Invest in waste management, redesign packaging, and enforce producer responsibility. 

- Funding for circular systems, not just gadgets. Technologies that trap plastic harmlessly near river mouths are promising, but they must be integrated into recycling and upcycling infrastructure. 

- Scaling community action. Over 48,800 volunteers in the Philippines alone collected more than 200 metric tons of trash during a single coastal cleanup day in 2025. That is the kind of mobilisation we need everywhere, supported by, but not replaced by AI.

If this is something that matters to you, please explore our Marine Litter project and follow the progress of this open environmental initiative:  at Marine Litter project (https://www.mi4people.org/en/marinelitter), or on our social media pages.

And if you care about what we do at MI4People with honest, practical AI for good, share this newsletter with your network to spark an honest conversation about tech in conservation. 

Or volunteer with us to help scale our open‑source models or build many more impactful solutions. Because together, we can build technology as a catalyst for change.