Source Cooperative

📢 The #GlobalBuildingAtlas rollout just leveled up! Last week I teased the arrival of the 2.75 billion features #GlobalBuildingsAtlas dataset to gee-community-catalog.org and it's finally complete. I have also added a brand new app with which you can explore which tile represents your data, download them in #parquet from #SourceCoop locally and much more. 👇 App: https://lnkd.in/gV7y3DRy 👇 Dataset Page: https://lnkd.in/g4nJC3JM 👇 Changelog: https://lnkd.in/gMJMDnnB Earth Engine applies it own compression to get the Vector dataset to about 250 GB from 1Tb but more importantly this work stands on the shoulders of many who believe true access means turning open data into usable, discoverable resources for everyone in the #geospatial community. If you like more updates like this and more help support the project and star out #Github repo https://lnkd.in/dR2QBrA and #subscribe to our newsletter at datacommons.substack.com #OpenScience #3DMapping #AI #AIforGood #BigData #ClimateAction #CommunityCatalog #CommunityTools #DataScience #DeepLearning #EarthEngine #EarthObservation #Geospatial #GeospatialAI #MachineLearning #ML #OpenData #OpenSource #PlanetScope #RemoteSensing #SatelliteData #SDG11 #SDGs #Sustainability #TUM #UrbanAnalytics #UrbanPlanning Emil Cherrington, PhD Xiaoxiang Zhu Ujaval Gandhi

Case Study: Bridges to Prosperity on Source Cooperative How do you address rural isolation, one of the most persistent drivers of poverty worldwide? In this case study, see how Bridges to Prosperity is tackling the challenge with data and #AI: ▪️ Built an impact model to predict how bridges transform access to health, education, and economic opportunities ▪️ Found that only 60% of their bridges aligned with mapped waterways ▪️Developed an #AI algorithm that tripled global mapped waterways (+124M km) ▪️Created a Waterways Explorer tool to visualize and share these insights The dataset is hosted on Source as PM Tiles, a choice that has already unlocked: ▫️ New funding ▫️ Greater credibility for digital approaches to infrastructure planning ▫️ Strong collaborations ▫️ Data use by governments, NGOs, and climate-resilience groups 📖 Read the full case study and share your thoughts!

🎙️ Source Cooperative's inaugural podcast is on Sep 30 at 11am PT! Join Jed Sundwall for an in-depth conversation with Drew Breunig, author of the blog post, "Why LLM Advancements Have Slowed: The Low-Hanging Fruit Has Been Eaten." They'll talk about the data products that fueled AI’s advancement and why they may have reached their limits. #LLM #AI You can join the conversation live via YouTube.

🪐 We believe that foundational planetary data should be open and accessible to all. Thanks to Source Cooperative, we’re able to share our datasets with the broader open data community. In line with our commitment to openness and collaboration, we’re excited to share what we’re learning along the way and to connect with individuals and groups from around the world. "Source has allowed us to open up all of this data to really anybody in the world… we would have been hesitant to do this without Source. We get people from local communities, indigenous groups, and others from almost every continent, who have reached out to us about this data." — Tom (Hutch) Ingold, Earth Genome, CTO 🔗 Learn more: https://lnkd.in/eiEbjqAF

We're starting a new podcast live stream thing! Every month, we're going to host a live streamed conversation about the craft and ergonomics of data. We'll record each episode and publish it as an audio-only podcast for people to listen to on their own time too. First episode is scheduled with Drew Breunig on Sep 30 at 11am Pacific. We're doing this because we believe the art of creating data products with great UX remains understudied and overlooked. It’s time to explore what sets great data products apart, looking at the history of data products that have had a huge impact on the world. Naturally, we're going to start with plenty of discussion about Common Crawl Foundation and ImageNet. Link to register in the comments!

A remarkable data product from Technical University of Munich made available in Parquet on Source. Link in comments.