---
title: "Mapping a smarter future with BigQuery and Google Earth AI models and datasets"
source_name: "Google Cloud Blog"
original_url: "https://cloud.google.com/blog/products/data-analytics/google-earth-ai-models-and-datasets-in-bigquery/"
canonical_url: "https://www.traeai.com/articles/2a82472f-4fdd-4092-b239-a7c49f65c588"
content_type: "article"
language: "英文"
score: 9
tags: ["BigQuery","Google Earth AI","地理空间分析","AI模型"]
published_at: "2026-04-27T16:00:00+00:00"
created_at: "2026-04-27T23:13:31.365095+00:00"
---

# Mapping a smarter future with BigQuery and Google Earth AI models and datasets

Canonical URL: https://www.traeai.com/articles/2a82472f-4fdd-4092-b239-a7c49f65c588
Original source: https://cloud.google.com/blog/products/data-analytics/google-earth-ai-models-and-datasets-in-bigquery/

## Summary

Google Cloud推出BigQuery与Google Earth AI结合的地理空间分析工具，包括街景、航空和卫星数据模型，助力企业实现智能决策。

## Key Takeaways

- 新增LiDAR数据支持，可精确测量基础设施高度和尺寸。
- 航空和卫星模型可通过提示词快速识别大规模地理对象。
- Vantor利用这些模型将原始卫星图像转化为灾后修复的可行洞察。

## Content

Title: Google Earth AI models and datasets in BigQuery

URL Source: http://cloud.google.com/blog/products/data-analytics/google-earth-ai-models-and-datasets-in-bigquery/

Published Time: 2026-04-27

Markdown Content:
Last year we [introduced](https://cloud.google.com/blog/topics/sustainability/new-geospatial-datasets-in-bigquery) new geospatial analytics capabilities integrated for BigQuery. Building on this, we announced an **expanded suite of tools** at Google Cloud Next ‘26, designed to help your business unlock deeper insights and make smarter, data-driven decisions. These Google Maps Platform models and datasets, leveraging innovation from [Google Earth AI](https://ai.google/earth-ai/), are integrated with BigQuery and Gemini Enterprise Agent Platform. They help you transform geospatial information into actionable intelligence, empowering you to understand our planet and its communities like never before.

### Harnessing AI for planetary understanding

In March [we launched Street View Insights](https://mapsplatform.google.com/resources/blog/turning-280-billion-images-into-actionable-infrastructure-insights-street-view-insights-is-now-generally-available/) in general availability, which draws on Google Street View’s vast repository of over 280 billion images and turns them into actionable understanding of physical infrastructure. This enables customers in telecom, utilities and the public sector to reduce weeks of manual work to minutes and get insights right from their desks. In the coming weeks we’re bringing the experimental release of LiDAR data to Street View Insights, providing precise measurements of infrastructure. With this, you can accurately determine the height of utility poles, the clearance of overhead lines, or the specific dimensions of road signs without having to manually gather measurements from the field.

We’re also expanding our Imagery portfolio in the coming weeks to include the experimental release of Aerial and Satellite Insights, providing a multi-perspective view of infrastructure that includes aerial, satellite and Street View imagery. This will help organizations manage assets at scale and with context. You can now combine top-down aerial and satellite views for large-scale planning and regional assessments with the ground-level detail of Street View to verify specific asset conditions.

Finally, we’re taking geospatial analysis to new heights with our Aerial and Satellite Models, developed as part of Google Research’s [Remote Sensing Foundation](https://research.google/blog/google-earth-ai-unlocking-geospatial-insights-with-foundation-models-and-cross-modal-reasoning/) effort, and now available in experimental within Model Garden. Now you can license our stand-alone model to build custom applications on any high-resolution aerial or satellite imagery source. Read our [blog](https://mapsplatform.google.com/resources/blog/unlocking-a-new-dimension-of-understanding-advanced-geospatial-ai-using-google-imagery) to learn more about Street View Insights, Aerial and Satellite Insights, and Aerial and Satellite Models.

![Image 1: https://storage.googleapis.com/gweb-cloudblog-publish/images/1_Fiq2gO6.max-1100x1100.jpg](https://storage.googleapis.com/gweb-cloudblog-publish/images/1_Fiq2gO6.max-1100x1100.jpg)

With Aerial and Satellite Models, an energy analyst can type a prompt like “find large HVAC cooling towers”. The model identifies relevant cooling tower objects across large geographies.

### How Vantor is using Aerial and Satellite Models

Following a severe storm, recovery teams need a clear picture of the damage to help communities rebuild. Vantor, a leading spatial intelligence company, uses these models in its Sentry application to turn raw satellite imagery into actionable insights. This helps organizations quickly identify washed-out roads and damaged infrastructure, so they can proactively remove storm debris and prioritize long-term repairs.

“The combination of Vantor’s spatial foundation and Google’s Aerial and Satellite Models is creating a new class of geospatial intelligence systems that can interpret activity across the planet, surface meaningful signals, and deliver insights directly into operational workflows. In demonstrations with customers, where we’ve integrated models into our persistent monitoring application called Sentry, the level of insight has been remarkable.” - Peter Wilczynski, Chief Product Officer, Vantor

![Image 2: https://storage.googleapis.com/gweb-cloudblog-publish/original_images/2_xPhNhU8.gif](https://storage.googleapis.com/gweb-cloudblog-publish/original_images/2_xPhNhU8.gif)

Vantor’s Sentry application uses Aerial and Satellite Models to turn raw imagery into actionable insights. After a storm, this helps their own users quickly identify washed-out roads and damaged infrastructure, so they can proactively remove storm debris and prioritize long-term repairs.

### Understanding communities

To learn about populations and their behaviors, researchers typically rely on three types of data sources — censuses, surveys, and satellite imagery — all of which are infrequently updated and can lack scale.

To address this, we’re announcing the preview of Population Dynamics Insights, a first-of-its-kind geospatial embeddings dataset powered by Google Research’s [Population Dynamics Foundation Model](https://research.google/blog/insights-into-population-dynamics-a-foundation-model-for-geospatial-inference/) ([PDFM](https://arxiv.org/pdf/2411.07207)) designed to help organizations decode the complex relationship between human behavior and the physical world. By distilling anonymized trends derived from Google search trends, Google Maps points of interest, busyness, air quality and pollen data into rich 330-dimensional vectors for places across the globe, it enables a new era of spatial machine learning without the need for manual feature engineering. Learn more in our [blog](https://mapsplatform.google.com/resources/blog/from-static-maps-to-geospatial-ai-announcing-population-dynamics-insights).

![Image 3: https://storage.googleapis.com/gweb-cloudblog-publish/images/3_VFBoi0e.max-1600x1600.png](https://storage.googleapis.com/gweb-cloudblog-publish/images/3_VFBoi0e.max-1600x1600.png)

### Safer and smarter road networks

We want to help local authorities make roads safer and smoother for everyone. That’s why we’re adding new preview features to Road Management Insights. You can now measure vehicle counts, to provide accurate traffic estimates that are required to evaluate the impact of new roads, bridges, and major maintenance projects. We’re also adding real-time disruptions for things like road closures that provide early signals about the potential reasons for traffic slowdowns. Finally, we’re announcing that Road Management Insights is expanding beyond the public sector, and is now available to logistics and roadside assistance companies. Get more information in our [blog](https://mapsplatform.google.com/resources/blog/roads-management-insights-expands-with-new-capabilities-for-the-public-and-private-sectors).

### Accelerate renewable energy adoption

We’re also introducing the experimental release of Solar Insights, now available in BigQuery. Built on the same imagery data and models available within Aerial and Satellite Insights and Aerial and Satellite Models, it provides high-resolution, building-level data on solar potential and existing arrays to help utilities and service providers accelerate renewable-energy adoption and optimize network planning. With Solar Insights, you can predict the next frontier of renewable energy market opportunities with BigQuery. Overlay information about solar potential per building, along with existing solar deployments to reveal untapped market opportunities and optimize investment strategies. Additionally, integrating these building-level details with our weather models and historical weather data allows you to accurately predict rooftop solar power contributions, increasing energy reliability and driving more profitable investments in renewable infrastructure. Learn more about Solar Insights [here](https://mapsplatform.google.com/resources/blog/unlocking-a-new-dimension-of-understanding-advanced-geospatial-ai-using-google-imagery).

### Optimize health and well-being

Understanding how environmental factors impact health is more crucial than ever. We're excited to announce new environment datasets, now available in experimental through Google Maps Platform. These datasets provide air quality, pollen and weather insights, and enable you to go beyond real-time data to unlock environmental understanding through hyper-local, high-resolution historical data that’s tightly integrated with BigQuery. This makes it easy to spot long-term patterns, like how allergy seasons affect your business or where air quality impacts public health. By mixing this environmental data with your own records, you can stop reacting to the weather and start planning for it. Whether you're deciding where to send resources or how to protect your customers, you’ll have the full picture of how the environment shapes your world. Read more in our [blog](https://mapsplatform.google.com/resources/blog/from-reaction-to-resilience-empowering-industries-with-advanced-environmental-intelligence).

![Image 4: https://storage.googleapis.com/gweb-cloudblog-publish/images/4_UUdqdpT.max-1500x1500.png](https://storage.googleapis.com/gweb-cloudblog-publish/images/4_UUdqdpT.max-1500x1500.png)

Visualizing the median PM2.5 levels in Manhattan on a specific day

### How these datasets can work together

One example of how these datasets can work together is Google for Health's [Population Health AI (PHAI)](https://blog.google/innovation-and-ai/technology/health/google-ai-heart-health-australia/), an advanced analytics eng ine that helps identify hidden health risks within communities. The goal is to equip our partners with insights that could help them shift from treating problems to proactively managing chronic condition risks. To provide this comprehensive view, PHAI utilizes Google Maps Platform’s Population Dynamics Insights, Places Insights and air quality and pollen datasets. By analyzing these diverse, de-identified data sets — ranging from geographic factors like the air we breathe to local access to fresh food — the AI model helps healthcare providers understand the shift from reactive treatment to proactive, tailored management of chronic condition risks for specific towns or postcodes.

Ready to explore what's possible? [Visit our website](https://mapsplatform.google.com/maps-products/geospatial-analytics/?utm_source=product-page&utm_medium=blog&utm_campaign=cloud-next-2026&utm_content=gmp-cloud-blog-website) to discover how Google's geospatial analytics can help you unlock your next big opportunity, or [sign up](https://mapsplatform.google.com/lp/geospatial-analytics-signup/?utm_source=landing-page&utm_medium=blog&utm_campaign=cloud-next-2026&utm_content=gmp-cloud-blog-signup) for early access and to learn more.

Posted in
*   [Data Analytics](https://cloud.google.com/blog/products/data-analytics)
*   [Maps & Geospatial](https://cloud.google.com/blog/topics/maps-geospatial)