AI + Environment Summit Zurich

🌍 Have you heard of “30 by 30”? At the AI + Environment Summit Zurich 2025 at Switzerland Innovation Park Zurich, we learned about this global initiative to protect 30% of our planet’s land and oceans by 2030, aiming to halt biodiversity loss, safeguard ecosystems, and fight climate change. But here’s the challenge 🤔 As Sara Beery (MIT) and Melissa Chapman (ETH Zürich) explained at last week’s AI + Environment Summit 2025: How do we decide which areas matter most to protect? 💡 With AI reshaping conservation efforts, data plays a huge role, yet wealthier regions often have more data, meaning models could potentially overvalue those areas and undervalue places with less information. Sara Beery’s research tackles another gap: Aerial data misses microhabitats like what’s happening under the forest canopy. By combining satellite data with ground-level imagery (like camera trap photos), her team captures and combines more details into new models. Meanwhile, Melissa Chapman focuses on translating massive datasets into fair, effective policy strategies, ensuring AI supports equitable environmental decisions. ✨ Want to dive deeper? Check out our YouTube playlist for all main stage keynotes from the AI + Environment Summit Zurich 2025: https://lnkd.in/dJ5XtNPZ

Most headlines about AI and the environment focus on heavy energy and water use. This week we saw something different: AI used deliberately to tackle environmental problems. At the AI + Environment Summit Zurich in the event hangar at Switzerland Innovation Park Zurich, the room was full of young, motivated people applying AI where it matters. In talks and posters, a clear pattern emerged: - Turning diverse, noisy, imperfect data into usable insight. - Biodiversity monitoring with computer vision. - Sensor fusion across satellites, drones, and ground stations for climate and ecosystem observation. - Restoring degraded imagery to enable long-term Earth monitoring. - Applied work in renewable-energy forecasting, sustainable agriculture, water and soil management, and the circular economy - using ML to optimize resource flows and anticipate impacts. Taken together, this points to an evolving ecosystem: maturing AI methods, expanding application fields—from biodiversity to energy and urbanization- and a shared understanding that real impact depends on interdisciplinary collaboration and responsible deployment. Used well and in dialogue with society, AI can help accelerate the transition to a climate-resilient and more just future. Thank you to everyone who contributed - speakers, poster authors, partners, and volunteers - for a day of focused talks and open discussion. Stay tuned: We’ll aim to share main-stage recordings soon. In the meantime let us know your most important take away or eye-opening moment from the talks and discussions! ETH AI Center, ETH Sustainability, ETH RAUM - The Platform for Sustainable Spatial Development, Zurich AI Festival, UZH Innovation Hub. Greater Zurich Area, Innovation Zurich

𝗦𝗰𝗮𝗹𝗲 𝘂𝗽 𝘄𝗶𝘁𝗵 𝗲𝘃𝗶𝗱𝗲𝗻𝗰𝗲: 𝗲𝗺𝗶𝘀𝘀𝗶𝗼𝗻𝘀 𝗰𝘂𝘁, 𝗵𝗮𝗯𝗶𝘁𝗮𝘁𝘀 𝗽𝗿𝗼𝘁𝗲𝗰𝘁𝗲𝗱, 𝗿𝗶𝘀𝗸𝘀 𝗿𝗲𝗱𝘂𝗰𝗲𝗱—𝗺𝗲𝗮𝘀𝘂𝗿𝗲𝗱 𝗮𝗻𝗱 𝗿𝗲𝗽𝗲𝗮𝘁𝗮𝗯𝗹𝗲. On Wednesday, the 𝗔𝗜 𝗘𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁𝗮𝗹 𝗦𝘂𝗺𝗺𝗶𝘁 took place at Innovation Park Zurich. Four keynote contributions stood out: 𝗣𝗿𝗼𝗳. 𝗗𝗿. Sara Beery (MIT) – AI & Biodiversity How computer vision can measure biodiversity at global scale despite domain shifts, incomplete data, and fine-grained taxonomies. Open question: How do we ensure data quality and representativeness across regions and species? 𝗣𝗿𝗼𝗳. 𝗗𝗿. Devis Tuia (EPFL) – Machine Learning for Earth From smartphones to satellites: translating multi-sensor data into robust environmental indicators. Open question: Which KPIs truly matter for cities and regions—emissions, land-use efficiency, resilience, or a composite index? 𝗗𝗿. Mikolaj Czerkawski (Asterisk Labs) – On the Evolutions of the EO Spheres, Data-centric pipelines, curated datasets, and generative models for Earth observation in practice. Open question: How do we build reusable, open datasets that scale across use cases? 𝗣𝗿𝗼𝗳. 𝗗𝗿. Xiaoxiang Zhu (TUM) – AI4EO for Social Good AI4EO at scale—from global urbanization to the SDGs—bridging research and real-world impact. Open question: Which governance models accelerate translation into municipal and regional planning? The summit underscored a shared agenda: making environmental intelligence reliable, interoperable, and actionable. Across biodiversity monitoring, urban KPIs, data-centric EO methods, and governance, the key challenge is moving from promising AI prototypes to standardized, open, and policy-ready tools for cities and regions. We must take into account the energy requirements of AI and limit them as much as possible. Thank you to the Team of AI + Environment Summit Zurich! You made it possible, to get a very interesting and perfect organised summit! ETH Sustainability, ETH RAUM - The Platform for Sustainable Spatial Development, ETH AI Center, Switzerland Innovation Park Zurich, Zurich AI Festival on Youtube: https://lnkd.in/dPCFaRGR more information: https://lnkd.in/d3dQWeDc