Spatial on Saturdays No. 7
The best of the week (AKA one big link roundup. No more, no less.)
By the numbers
1,700,000
Electric vehicles on the road in the United States today (source)
26,000,000
Electric vehicles projected to be on the road in the US in 2030 (source)
~1,800,000,000
Buildings in the Google Open Buildings Dataset focusing on Africa, South Asia, South-East Asia, Latin America and the Caribbean
Best of the week
For some reason, this week was a busy one for some great content around the web. So here they are.
🏠 ICYMI, Google Open Buildings. Google released a massive dataset of buildings interpreted from imagery with about 1.8 billion polygons in Africa, South Asia, South-East Asia, Latin America, and the Caribbean.
🗻 Speaking of Google. This post from Pavel Nikolov shows the combination of the recently released 3D photorealistic map tiles from Google Maps and Cesium 3D which produces a stunning result of an almost real 3D flyover.
Geospatial @ Coursera
Check out some of my geo-focused favorites from Coursera:
Geographic Information Systems (GIS) Specialization from UC Davis:
Spatial Data Science and Applications from Yonsei University:
Python for Everybody from University of Michigan
Remote Sensing Image Acquisition, Analysis and Applications
📍Why quality international point of interest data is hard to come by. POI data is a key part of many applications, base maps, analyses, and business datasets. Google Maps has often been the standard of this data, but it can’t be accessed for batch analytics and still often lacks accuracy in various countries. The team at dataplor, who collects this data using machine learning, call bots, local experts, and image analysis shared a great post on why this data is so hard to collect accurately, and the potential ramifications.
🎉 Congratulations are in order! Congrats to Dr. Qiusheng Wu on passing 1 million views on his YouTube account. With 510 videos shared over 3 years, that is about a video every 2.15 days.
🚰 Hydrology tutorials using Whitebox. One tool I have been following is Whitebox tools which allow you to perform a number of raster analyses, and it is also integrated into leafmap. This tutorial walks through a hydrological example and I hope there are more to come!
🛣️ A roadmap for geospatial and generative AI? Matthew Lewin from Esri Canada shared a roadmap for generative AI and geospatial technology in different buckets such as data acquisition, validation, data discovery, analysis/modeling, visualization, and sharing/collaboration.
Yes, but. I think this vision may develop but some things are a ways from realization such as image enhancement, automated data QA, complex feature detection (apart from things like street view images), and high-quality map generation.
Some are already happening. That being said, things like data augmentation via labeling, intelligent data search, synthetic data creation, and data anonymization are already happening in the geospatial and broader AI fields. Once more tools are built to connect the AI models to structured data, these tasks will become even easier.
🎙️ OpenStreetMap GPT: One practical use case for generative AI that is already happening in several places is with querying data using text inputs, and Rohit Gautam put together a new project that allows you to query the OSM Overpass API with just text, and it looks amazing. Follow him for the open-source release coming this week/
🖥️ Webinars. And more webinars. Make sure to check out these two virtual events:
Modern GIS and Data Visualization with Helen McKenzie
7/27 @ 1p ET - Spatial Analysis in Snowflake with Fawad A. Qureshi (Field CTO @ Snowflake) and Me
🔌 Isochrones and electric vehicles. First is a very cool Streamlit app from Arthur Dolmajian that allows you to test different techniques to create isochrones (article, post, and app) from OSM data. And this post from Pieter Waller, CCO at Chargetrip which is a range predictor for EVs, shows exactly why isochones matter in the new world of EVs.
Thanks!
If you have something you want to share or a job you want to promote you can e-mail me here. Or if you want to sponsor a newsletter, video, or post, check out this link to get in touch.