SAS’ mission is to empower and inspire with the most trusted analytics, helping uncover hidden insights that break down barriers, fuel ambitions and get results. Every year we find innovative ways to use advanced analytic technology to solve pressing humanitarian issues and make a positive difference for people and our planet.
Rainforests continued to experience devastating forest loss as destruction and deforestation levels spiked while conservation and enforcements efforts lagged due to the pandemic’s ripple effect. The Amazon is the world’s largest rainforest and homes an impressive 10% of the world’s biodiversity. Unfortunately, deforestation is wreaking havoc on the Amazon; approximately 800 square kilometers of forest is destroyed every month. Even worse, during the COVID-19 pandemic, there’s been a surge of deforestation in Brazil. This environmental crisis is no small feat – and we knew we would need all hands on deck to help drive change.
SAS launched a global social innovation project to use crowd-driven artificial intelligence to help track, and ultimately stop, deforestation in the Amazon rainforest. We asked an army of citizen data scientists to review satellite images of the Amazon in order to better train existing AI models detecting deforestation. Users classify if there are signs of human impact in the images, helping improve the AI algorithm and expedite the analyzation process. The plan was to gain an improved view of deforestation to help support critical policies to quickly and effectively protect forests.
On Earth Day 2020, we announced a partnership with the International Institute for Applied Systems Analysis (IIASA), an independent research institute that provides baseline information and policy solutions on pressing concerns for humanity. IIASA research projects are multiscale and interdisciplinary, spanning a wealth of environmental and sustainability topics. We leaned on IIASA’s expertise and thorough research in deforestation to ensure our application was effective and that the outcomes couldbe used to drive real, tangible change. We targeted the Amazon for this effort based on existing research.
While computer vision models can be trained to quickly identify areas of the rainforest that have been significantly damaged, signs of deforestation can be challenging for a computer to see at first. It takes human eyes to properly classify images in order to build models that can detect the subtle differences between satellite imagery.
From Earth Day on April 22, 2020, through February 2021, citizen scientists classified nearly 90,000 satellite images of the Amazon rainforest in phase one of the project. Every image that volunteers selected as an area of deforestation in our crowdsourcing project got SAS and our partners closer to building a model that can alert governments and conservation organizations.
During the second phase, which started March 2021, we continued combining SAS AI technology, human input and our partner’s expertise to identify where changes are occurring over time. This could one day help predict where deforestation is likely to happen next.
This project is unique due to its approach to incorporate participation and perspectives from a variety of participants to combat the issue of deforestation. Our innovative work with IIASA uses our trusted AI and computer vision technology and the power of the crowd to propel us closer to real change – one image at a time. Not only does tapping into collective knowledge help us achieve results faster, which is critical given the current unprecedented rate of deforestation and habitat destruction, but it also helps show the world the human side of AI.
All of our communication and promotion efforts point people back to our microsite, encouraging them to spend some time identifying deforestation in the satellite images shown. And we are transparent about the process – from how the images are collected to how the models are evaluated and deployed – providing simple explanations behind the often-misunderstood technologies.
This work demonstrates the impact of crowdsourcing, which will be incredibly important in tackling societal problems of this size. By having average citizens assist with this analytic work, we are showing them the behind-the-scenes work of AI and visualization technology. We gain a larger pool of data to analyze, while simultaneously humanizing technology to the general public. In order for our industry to keep moving forward, we need people to understand and trust the technology that will help drive real change for our society. By giving people a seat at the analytics table, they can witness the magic behind AI for themselves. This can also help equip the next generation with the data literacy skills they need to succeed in our digital world moving forward.
We launched this project on Earth Day, taking advantage of timely global conversations and celebrations about sustainability and conservation. With the COVID-19 pandemic sweeping the globe during this time, many in-person Earth Day activities were canceled, leaving people feeling like they were unable to make a strong impact on the annual holiday. Our project, however, allowed people of all ages to make a measurable impact on our planet. Our execution included traditional media, web, social media, advertising and online events. Resulting media coverage about the project led to a significant spike in user participation. The most exciting result was that the average visitor sorted 49 images.
The project launched with more than 44,000 satellite images and now has successfully classified 1,047,996 square kilometers of the Amazon as of March 2022 when that phase of the project was completed. Participants joined the initiative from 130 different countries, highlighting the project’s global scale. In 2021, this crowdsourced data helped to train the AI model and data scientists subsequently validated its efficacy in identifying deforestation. SAS will be publishing a joint research paper with IIASA about this important new approach to conservation in 2022.
SAS has furthered its commitment to making a difference by joining forces with the nonprofit Amazon Conservation. This project will expand the scope and efforts for identifying and tracking illegal deforestation and expediting intervention by monitoring key parts of the Amazon. These findings also provide opportunities for innovation not just for Amazon Conservation but hopefully beyond the Amazon to address environmental issues globally.