At the beginning of this month, the Google Research team introduced Geospatial Reasoning, a project that applies generative AI to the mapping and location search industry. The challenge of locating precise geographic data is a complex task that requires high expertise and extensive manual work. Integrating AI into this process can significantly streamline the workflow.
Geospatial Reasoning is a component of Gemini that analyzes data in collaboration with various specialized geographic models, such as the Population Dynamics Foundation Model (PDFM) which links population behaviors with the environment, trajectory-based mobility foundation model that studies population movement, and remote sensing foundation models for additional insights.
The outcome enables us to input prompts and complex tasks for the models to assist with. For example, asking “how many medical buildings were damaged in the recent hurricane?” The system can then analyze this data and provide answers on the number of damaged buildings, hospitals, clinics, their locations, and postal codes.
Currently, Geospatial Reasoning is in its prototype phase, resembling the interface of Google Maps or Google Earth, but with the capability to interact with Gemini. It is currently accessible for testing by selected trusted partners only.
Source: Google Research
TLDR: Google Research introduced Geospatial Reasoning, empowering mapping and location search with AI technology. The system collaborates with specialized geographic models to streamline complex data analysis tasks, providing valuable insights for users through a user-friendly interface.
Leave a Comment