Header image

How Computer Vision Can Help Fight Forest Loss

March 21, 2022 blog-post artificial-intelligence big-data climate-action climate-change computer-vision environment geospatial machine-learning open-data partnership remote-sensing satellite-imagery southeast-asia sustainability development geospatial ngo non-profit

We are excited to announce that this year, Thinking Machines is working with the Gerry Roxas Foundation to use earth observation data and machine learning methods to map the drivers of deforestation in the Philippines.

A Global Forest Crisis

Ecosystems around the world are under threat, with deforestation and biodiversity loss occurring at an unprecedented rate. Southeast Asia alone, one of the world’s largest carbon sinks, has already lost more than half of its original forest cover in the past two decades (Earth.org). This is why the United Nations has declared the years between 2021 and 2030 as the “Decade on Ecosystem Restoration,” which aims to build a global movement to halt the degradation of ecosystems to enhance livelihoods, counteract climate change, and halt the collapse of biodiversity.

The Gerry Roxas Foundation (GRF), one of the oldest foundations in the Philippines, is working to address deforestation in hotspots across the Philippines. Over the next 5 years, GRF will administer $16M in grants to civil society organizations working to restore ecosystems in 30 key bioregions of the Philippines. To use this funding wisely, GRF needs comprehensive, up-to-date information on the key drivers of forest loss in its areas of concern. Unfortunately, government maps on forest cover and land use are only updated every 5 to 10 years. The last forest cover maps produced by the DENR were released in 2020 and based on data collected in 2015. And although open data portals like Global Forest Watch provide useful global data, they lack the detail and nuance that GRF and their partners need to make decisions at the local level.

Supporting Ecosystem Restoration with AI-Generated Data

This is where AI can help. By applying computer vision to the latest satellite imagery, Thinking Machines will work with GRF to produce detailed maps of how forest cover and land use have changed over the past decade in GRF’s areas of concern. Our goal is to develop a fast, automated, scalable approach of producing data to complement and augment ground-truth provided by the local organizations receiving grants from GRF.

Sample output of deforestation driver classifications for an area in Indonesia, from the methodology of Irvin et al., 2020

Armed with this data, GRF and its grantees can make the biggest possible impact with their available resources.

"When there is no time to lose, we cannot keep doing things the same way and expect our conservation work to succeed. We must be guided by data to make strategic decisions on investment in protection and restoration," says Ipat Luna, Chief of Party for GRF's ecosystem restoration grant program.

We will share the results of our collaboration with GRF in the coming months.

Thinking Machines’ Eco-Intelligence Solution applies end-to-end analysis to a robust data catalog of curated and AI-derived environmental datasets to help clients like GRF with data they need to accelerate ecosystem restoration efforts. To learn more about this solution, contact us at [email protected]

Photo Credit: Bill Ringer via Unsplash, hiking trail in Mt. Ugo, Philippines


Quarantine in a Multicultural Hub: Exploring Mapbox Movement in Singapore

Successful businesses use geospatial analytics wisely and we’ll tell you why you should start using it too.

Zero in on the Philippines’ most vulnerable communities – with the click of a mouse

Estimate wealth and poverty for any 18 square kilometer area within a fraction of the time and cost of running a household survey.

Send Your Friends a Digital Holiday Card With Our AI Holiday Greeter!

Generate your own digital holiday greeting card with artificial intelligence and beautiful data visualization and share it with your loved ones this holiday season!