Superlinear’s AI for Sustainability series
In this article, you’ll discover the insights provided by Victor Hutse, Solution Lead at Superlinear, who shared his thoughts on the role of AI and, more concretely, Machine Learning in conservation, forests, and farms.
Watch his presentation below, or continue reading instead ⬇️
The role of AI in conservation today
The relationship between conservation and climate change mitigation is strong. The world's forests capture 7.6 billion tons of CO2 annually, equivalent to 300% of European or 150% of the US yearly emissions. Using AI as a copilot, we can leverage both data and existing expert knowledge on conservation. This will allow professionals to access those insights to make more informed decisions efficiently.
Let’s dive into several use cases to learn more.
Ocean level
Monitoring Endangered Plants: By training Machine Learning models to recognize endangered plants on drone images, large areas can be surveyed, creating maps displaying the locations of these plants. This approach allows for better monitoring and conservation of endangered species. Citizen scientists can fly over large areas of fields, recognize the plants, stitch images together, and recreate maps for detailed analysis.
Tracking Whale Populations: Machine Learning can help monitor whale populations by scanning satellite images, detecting possible whale locations, and identifying specific whales. This information can be used to create interactive maps for scientists and policymakers to track whale populations. The approach involves training a model to scan ocean images in large parts, detecting patches where whales might be, and another model looking at a closer level of detail to identify specific whales.
Ground level
Monitoring Wildlife with Camera Traps: Machine Learning can analyze images captured by motion-sensor camera traps in forests, identifying the species present and providing valuable insights into animal populations. Using microphones in conjunction with images allows for monitoring larger patches of space and time. This method can detect the movement of animals like elephants through the forest and also detect poachers.
Example. Detecting Animal and Poacher Behavior: The PAWS project by Harvard University uses machine learning to predict animal and poacher behavior, optimizing the routes of park rangers to increase their chances of encountering animals or poachers. This information is provided to park rangers through a dashboard or a map with suggested routes.
Micro level
Example. Identifying Animals and Plants with ObsIdentify: This app allows users to take pictures of plants and animals in the wild and uses machine learning to identify them. The data collected helps scientists understand species distribution and contributes to conservation research. The app is made by Observation International and is related to Natuurpunt in Belgium. The resulting database, observation.org or waarnemingen.be, provides a valuable resource for research on animal and plant distribution.
Example. World Forest ID: This organization collects wood samples and creates a fingerprint database to help customs officials identify and flag illegally sourced wood or other products, such as soy or cocoa. This effort aids in the fight against illegal deforestation. The organization is based in the UK and the US, working with several universities. Wood samples are collected from various forests and locations, and analyses are conducted, including microscopical and chemical reaction tests, to construct fingerprints.
In conclusion
Conservation plays a vital role in reducing net emissions and mitigating climate change. Artificial Intelligence has the potential to significantly impact conservation efforts by reducing the prohibitive costs associated with these initiatives. The various use cases above show that AI can provide scalable, innovative solutions for monitoring and protecting endangered species, ecosystems, and habitats. By harnessing the power of AI, we can continue to advance conservation efforts and work towards a more sustainable future.
Are you ready to welcome AI as your copilot? Don't hesitate to reach out to Superlinear!
