2018 & 2019 WinnerAn integrated data pipeline for small-scale fisheries
The Inspire Challenge is an initiative to challenge partners, universities, and others to use CGIAR data to create innovative pilot projects that will scale. We look for novel approaches that democratize data-driven insights to inform local, national, regional, and global policies and applications in agriculture and food security in real time; helping people–especially smallholder farmers and producers–to lead happier and healthier lives.
This proposal was selected as a 2018 pilot project and 2019 scale-up runner up project, with the team receiving a total of 225,000 USD to put their ideas into practice.
An integrated data pipeline for small-scale fisheries
Every day, about 40 million small-scale fishers go out fishing, yet virtually none of these activities or yields are documented. This global data deficiency underpins SDG 14, yet we are now at a moment in time when technologies are small enough and cheap enough to solve it. WorldFish are testing novel linkages of these technologies to create an automated data pipeline to highlight temporal and spatial changes in fish production. This pilot will provide a national proof of concept in Timor-Leste before transitioning to scale.
WorldFish will lead on-the-ground coordination of catch documentation and government partner capacity building and develop an open-source cloud-based decision dashboard for fisheries managers. Pelagic Data Systems (PDS) will be responsible for the installation of vessel-tracking hardware, transmission of geospatial tracking data through cellular networks, and the cloud-based analytics and machine learning necessary to identify fishing activity at sea.
The fish food system supports over three billion people globally and represents more than 20% of their animal protein. Approximately half of the global fish catch comes from small-scale fisheries, yet this contribution is hidden and small-scale fishers are often overlooked in policy, management, and development decisions. Ultimately, the marginalization of small-scale fishers undermines the vital food and livelihood benefits they provide.
The involvement and activities of women and youth in fish food systems are particularly under-reported. The existing data systems underpinning this project have followed CGIAR sex-disaggregated data standards which contribute to closing gender data gaps. This work has already highlighted the crucial importance of women in fishery activities as dependable sources of household food and nutrition security in Timor-Leste, and the approach is being deployed in other regions.
Wilderlab, a biomonitoring service based in Wellington, New Zealand, develops, improves, and maintains R scripted pipeline from catch data records to peskAAS decision dashboard. They also carry out statistical testing of nearest neighbour assignment.
WorldFish provides project leadership, on-the-ground activities, project implementation, and project reporting.
Shaun Wilkinson explains the data tool behind the initiative.
Step by step
Registration of fishers and installation of boat trackers
The WorldFish in-country team, in partnership with the Ministry of Agriculture and Fisheries (MAF), has deployed roughly 300 Pelagic Data Systems boat tracking units on artisanal fishing boats around Timor-Leste, including the enclave of Oecusse.
This partnership enables the project to support MAF in augmenting the national registration of small-scale fishing boats while installing tracking units at the same time.
Filtering data using machine learning
By teaching machines how to identify fishing activity based on the speed, trajectory, and tortuosity data collected by the PDS tracking units, fishing activities can be filtered from not-fishing activities and potentially flag for suspicious or unusual movement.
Capacity building, training, and participatory development of data systems
The project will provide three aspects of capacity building and training for government partners, fisher stakeholders, and data collectors:
- Installing and maintaining PDS trackers, attaching trackers to the boat in a way that is durable, non-damaging, and effectively exposing trackers to sunlight
- Collecting landing data
- Manipulating and comprehending decision dashboards
WorldFish and PDS trained and supported 11 new data collectors (one per coastal municipality) joining MAF.
A workshop with the Fisheries Directorate was held at the end of April to assess progress and evaluate how tools and processes match needs and aspirations.
Launch of near real-time monitoring system
The team launched the dashboard, Automated Analytics System for Small-Scale Fisheries in Timor-Leste (PeskAAS), which tracks fishing activities–including the number and type of fish caught by individual boats in near real-time.
Making headlines in Timor-Leste
The project made headlines in Timor-Leste for the launch of its near real-time small-scale fisheries monitoring system. The dashboard is one of the most sophisticated data collection systems for small-scale fisheries in the world. It puts important data in the hands of fisheries officers, researchers, and local stakeholders. The tool will ultimately enable better understanding the contribution of fish and fisheries to local livelihoods and food security.
Journal article publication
The first journal article using the peskAAS monitoring system was published in Frontiers in Marine Science.
The study tests the effects of fish aggregating devices (FADs) at increasing fish production in Timor-Leste. FADs have shown success in improving catch rates elsewhere in the Pacific by making pelagic stocks more accessible to small-scale fishers.
Eight experimental FADs were deployed at four sites around Timor-Leste and catch and effort data from FAD and non-FAD fishing trips was recorded. Results show a significant positive effect of FADs on catch rates when controlling for random site variation, with FADs paying for themselves in roughly five months or less at three out of four sites.
System testing in seven additional countries in Asia and Africa
The team signed an agreement to expand their pioneering work to better understand stocks and catches from small-scale fisheries across seven countries in Africa and other parts of Asia.
The Memorandum of Understanding, signed at a special ceremony at WorldFish headquarters in Penang, Malaysia, provides the framework for WorldFish and Pelagic Data Systems to further develop and provide access to innovative data collection solutions for fisheries.
“Working closely with innovative partners in the private sector – especially the tech sector – can help us generate more and better data, which in turn can mean our research has a greater impact in supporting some most vulnerable people in the world,” said Dr. Alex Tilley, Scientist at WorldFish.
Evaluation of uptake, sustainability, and scalability
In the final stages of the pilot phase, an evaluation workshop will provide a final assessment of the system’s capacity and impact, an evaluation of uptake, and an assessment of the system’s sustainability and scalability to other countries and fisheries.
US$125K scale-up grant
The project was awarded a 2019 Inspire Challenge Scale-up grant of US$125K at the third annual convention of the CGIAR Platform for Big Data in Agriculture, 16-18 October 2019.
Ongoing updates to the application dashboard
Updates are ongoing to the dashboard on two levels:
- Augmented (paid tier): re-imagining fishing data, testing with stakeholders, integrating new system, using German train-tracking model to look at how fish catch quantities relate to time spent fishing (efficiency analysis).
- Automated (open access, pending financial sustainability): optimizing for use by fishers
Updates are scheduled to be completed by August 2020.
Financial sustainability model report
The team is working on producing:
- A brief of the peskAAS system with which to approach impact investors
- A report detailing potential business models for the scaling of the peskAAS fisheries data system, including a brief strategy summary and at least two different scaling approaches.
Journal article review and publication
An article titled, “PeskAAS: Automated analytics of relative fish abundance over time and space” is in review, and publication is expected in March.
Addition to project team
A new data scientist from WorldFish will join the team.
Data security and ownership case study/manual
The importance of defining data security and ownership has become ever more apparent. Challenges faced by the team regarding defining their own accountability, inspired them to develop a manual on the subject, in collaboration with the BIG DATA Platform.
Field testing of components is planned to take place in Lake Malawi, Cox’s Bazar, Bangladesh, and Penang, Malaysia. These locations are already transmitting vessel movement data.
Two pelagic tracking units have also been sent to Egypt, Nigeria, and Zambia, and plans are in place to improve data yield in these locations.