AI, ocean data targets plastic for Indo-Pacific cleanup

By Mary Bolling

Plastic waste in oceans is choking coastal ecosystems – but a CQUniversity-led project is integrating live weather and ocean data with new artificial intelligence (AI) models to prevent trash hotspots in Jakarta Bay, Indonesia.

The NO-PLASTIC project funded by Commonwealth Scientific and Industrial Research Organisation (CSIRO) through its Indo-Pacific Plastics Innovation Network (IPPIN), has developed an AI-enabled system that can detect and map plastic pollution in real time. 

Teams from Australia, Singapore, and Indonesia are led by CQUniversity Senior Lecturer in Artificial Intelligence Dr Anwaar Ulhaq and Senior Research Fellow Dr Angela Capper, who recently visited the region to trial the technology.

“The NO-PLASTIC project uses new computer-vision models to analyse CCTV and drone footage from major waterways in Jakarta Bay, and predicts movement from live weather and ocean data,” Dr Ulhaq explained.

“Tides, urban runoff, and river flows all intersect to create recurring plastic hotspots, and our system means local teams can deploy trawl-nets and clean-up gear to the precise locations where plastic accumulates, improving the efficiency of river-to-sea clean-up operations.”

Global problem, local solutions

The United Nations’ Environment Programme estimates more than 11 million metric tonnes of plastic enter the world’s oceans every year, and Jakarta Bay is one of the Indo-Pacific’s most affected marine environments.

NO-PLASTIC international collaborators include Patrick Dupuis from Global MetOcean in Singapore, and Nicolas Bernier from OceanKita in Indonesia. 

The technology is designed to support local operators rather than replace them, offering a practical tool that strengthens decision-making and reduces the amount of plastic reaching wider marine ecosystems, including Australian shores.

“OceanKita’s unique local connections help to drive not only clean-up efforts but also community education, which is crucial to shift habits and attitudes about waste,” Dr Capper said. 

In November, Dr Ulhaq and Dr Capper visited Singapore and Indonesia to meet partners, evaluate the system at monitoring sites, and observe field conditions firsthand. 

Their work will refine how environmental data from Global MetOcean feeds into the AI model, and how OceanKita’s clean-up teams can use the system in busy coastal settings. 

The on-site work also ensured that the AI tools are adapted to local conditions, unstable weather, variable tides, and rapidly changing plastic movement patterns.

Collaborating for a cleaner future

The NO-PLASTIC project is model of Indo-Pacific collaboration, combining scientific expertise, local operational knowledge, and regional commitment to marine stewardship. 

Each partner plays a distinct role: Australia provides the AI, plastic pollution expertise and research design, Singapore contributes high-resolution environmental oceanographic and meteorological data, and Indonesia drives the practical deployment of pollution-reduction technologies. 

The Indonesian government has been working to reduce local marine plastic debris by up to 70 per cent by the end of this year, while CSIRO – Australia’s national science agency has a goal to see an 80 per cent drop in plastic waste entering the environment by 2030.

“By merging artificial intelligence with environmental science and community-led clean-up actions, our NO-PLASTIC project offers a scalable approach to identifying and reducing plastic pollution before it spreads,” Dr Ulhaq said.

“It’s positioned to inform policy, strengthen local capability, and contribute to healthier coastal ecosystems for communities across the region.” 

The team hopes the model can be adapted for other regions to tackle similar challenges, with smarter, technology-enabled marine stewardship.

Explore more projects driving smart and real-world solutions for people and planet at the CQUniversity Research website.