An important new report from The Alan Turing Institute’s report, commissioned by the Geospatial Commission, has been released. I wanted in this article to summarize the main points and highlight an important comment made by Ed Parsons of Google on the report
The report discusses the integration of geospatial AI into land use decision-making. Many see this as a groundbreaking opportunity. In many ways this report is a deep dive into how Geospatial 2.0 – leveraging new data sources, AI, and advanced platforms—can revolutionize urban planning, policy-making, and sustainability efforts.
For those who don’t have time to read the full report, here’s what you need to know.
Why Geospatial 2.0 Matters
Land is a finite resource, under immense pressure from population growth, net-zero goals, and the need for food and energy security. Traditional methods of land management rely on outdated, expensive, and infrequent datasets. Enter Geospatial 2.0—a new era where earth observation, advanced AI models, and dynamic data processing offer scalable, actionable solutions to make land use planning more informed and inclusive.
At the core of this revolution is geospatial AI, capable of transforming raw data into actionable insights. The report highlights how tools like the DemoLand prototype are pioneering ways to simulate land use scenarios, making data-driven trade-offs accessible to all stakeholders.
Key Insights and Use Cases
- Satellite Imagery: The Backbone of Geospatial AI – Satellites like Sentinel-2 generate frequent, low-cost data across vast regions, making it easier to track urban changes, monitor environmental impacts, and simulate future scenarios. With foundational models like SatlasNet, geospatial AI is unlocking predictions for air pollution, housing prices, and beyond.
- LLMs for Natural Language Accessibility – Large Language Models (LLMs), like ChatGPT, are lowering barriers for non-technical users by enabling conversational queries. Imagine asking, “How will air quality change near schools?” and receiving a plain-language answer backed by data. Tools like the DemoLand chatbot make this a reality.
- Combining Data for Better Insights – By integrating satellite imagery with traditional datasets, geospatial AI achieves nuanced insights. Adding geographic context further enhances accuracy, making it easier to capture regional disparities and patterns.
- Generative AI for Visualization – Generative AI can create visual “before-and-after” images of proposed land changes, helping non-experts grasp the implications of planning decisions.
Challenges and Solutions
The adoption of geospatial AI isn’t without hurdles. Key challenges include:
- High costs: Training and running foundational models require significant computational and data resources.
- Knowledge gaps: Combining expertise in AI, geography, and policy is rare but essential.
- Data accessibility: Licensing issues and reliance on closed-source AI models limit scalability.
The report recommends:
- Developing open-source geospatial AI toolkits.
- Increasing access to computational resources.
- Promoting interdisciplinary collaboration through workshops and shared platforms.
The Big Takeaway
This report isn’t just about land use planning—it’s about redefining how we understand and interact with the world around us. Geospatial AI is paving the way for more sustainable, inclusive, and data-driven decisions.
Yet, there is one critical challenge to address. As geospatial expert Ed Parsons notes:
“LLMs have little understanding of geography or the relationships between geographical entities within the study area. This makes it hard for them to interpret questions that involve reference to specific spatial features or areas.”
This is the next big hurdle for the geospatial industry to solve. Tackling it will unlock the full potential of Geospatial 2.0 and reshape the future of land use planning.
So, geospatial professionals, this is your call to action. The tools are evolving, the challenges are clear, and the opportunities are limitless.
Matt Sheehan is focused on the rapidly evolving geospatial world – Geospatial 2.0
Stay up to date with all things new in geospatial – Join the Geospatial 2.0 Community: https://spatialnext.io/subscribe-to-our-newsletter/