Liked what you read?

Subscribe to our monthly newsletter to receive the latest blogs, news and updates.

Join the Community

Introduction

“I want to make it 1,000 times cheaper to make the next Google Maps engine.” 

In the latest episode of The Ellipsis Drive Podcast, we hosted Jed Sundwall, Executive Director of Radiant Earth – a non-profit organization committed to increasing the understanding of our world through community-led initiatives that make spatial data seamless to access and use. 

As a team that promotes the growth of geospatial data accessibility ourselves, our CEO Rosalie was excited to talk with Jed and pick his brain about the current state of the geospatial ecosystem. This blog captures the core message of that conversation and sheds light on some of the key challenges, potential solutions, and technological trends awaiting the industry in 2025. 

So what would prevent us from making the next Google Map engine 1,000 times easier to create? Let’s find out!

Challenges in the Status Quo

In a rapidly changing and data-driven world, achieving a shared understanding of our planet is crucial. Geospatial data is the foundation for understanding patterns, relationships and changes across the Earth’s surface. Thus easy access to geospatial data holds immense potential value. 

But seamless accessibility to geospatial data is a rarity in the status quo. Below are two of the key reasons –

Data Complexity

Data is getting bigger. The need for significant compute power has grown. There are many more sensors around the world. It’s easier to launch satellites in space. There’s a lot more network activity as a result. All of these trends have created a highly complex data palette. A palette wherein each dataset has its unique infrastructural requirements, protocols, and tools.

“Data is too big to fit on a single device,” said Jed. “Data users have to get used to working with data stored remotely in cloud environments. It’s still a big paradigm shift for most people. One of our initiatives, Cloud-Native Geospatial (CNG) Forum, offers a vendor neutral platform for data practitioners to share their experiences and best practices with one another.” 

Such community-led initiatives lower the threshold for working with remote data sources, and thus, improves seamless accessibility to help people overcome some of their unique data complexity challenges.

Liked what you read?

Subscribe to our monthly newsletter to receive the latest blogs, news and updates.

Join the Community

Geospatial Data Silos

Geospatial data is often seen as a unique snowflake. A novelty item that requires its own unique infrastructure, expertise, and management. This causes geospatial data sets to become notoriously siloed. 

“The status quo is that you interact with geospatial data via a unique interface created solely to display geospatial data. That’s a big challenge.” said Jed. “A lot of business decisions are made in business intelligence systems that are not compatible with a map. Geospatial data can’t fit into those environments. We need to figure out how to break these GIS silos.”

Infrastructural Solutions for breaking down Geospatial Silos

Jed had a clear vision of how geospatial silos are likely to be broken down. “We have this notion that someone’s going to create a SaaS solution and it’s going to solve everything. That’s not going to happen. My vision is a world with many, many solutions. Solutions that are bespoke and esoteric. Solutions that increase accessibility to geospatial (environmental) data on a local scale. So that’s our overarching approach at Radiant Earth – to lower the cost of working with geospatial data so that local governments and businesses can come up with specific solutions for their areas.”

To enable the power of the aforementioned esoteric solutions, a solid foundation for data storage is crucial. Cloud based object storage platforms like Amazon S3, Azure blob storage, or Google Cloud Storage are affordable, flexible and commoditized tools that provide the base infrastructure. Anything hosted on these object storages can be churned out via https or another  preferred API.

Future Outlook

Table Engines vs Map Engines – 

“I’m very excited about columnar formats like GeoParquet.” said Jed. “There’s also a lot of work underway to make multidimensional array data easier to access in the cloud, like Icechunk.” Making geospatial data easier to work with in commodity cloud environments helps break down geospatial silos and makes data easier to integrate with business systems…We need to transcend ‘the map’ if we want that information to enter the decision maker’s mind.”

Jed tells us that he learned a lot from his experiences at Amazon. He shared how maps were always distilled down to a tabular format. And that tabular data was eventually condensed into a sentence or two, which was then used by a decision maker. 

“And while humans are visually biased beings, most data is likely to be consumed by machine algorithms or AI agents.”

Rosalie had a slightly different take on the matter. Environmental data has risen in importance due to volatility in climatic conditions and thus, on risk portfolios as a result. Most of these - climate and environment related - data points are natively raster data. And translating raster data into table engines means losing a lot of inherent spatial awareness.

“The need to develop a raster-engine exists when feeding the data into various types of AI models or running a complex query that needs to be spatially aware. Because you can’t condense a raster image and turn it into a table entry whilst not losing information on the nearness of spatially correlated assets. For example, to calculate the aggregated risk of flooding for a set of assets, you can’t ignore the need to distribute map-native data across the nodes in a compute cluster in a spatially aware way. That’s where I see a limitation of table-engines and an opportunity for raster-engines.” said Rosalie. “Not to understate the importance of being able to summarize raster data into a table. It’s just that the ability to do this effectively, is not the answer to every single one of our most crucial spatial data analysis challenges.”

Data Standards – 

“It’s really hard to create good standards. And they count only if they’re in use. Standards become obsolete if they’re too abstract or complex to implement.” said Jed.
In the current age of the internet, a handful of smart people could identify a pattern of data use and label it as a standard. Standards are useful when they attain a certain stature and size and OGC is the right organization to develop those standards.

Radiant Earth is specifically dedicated to helping communities find open, accessible patterns and best practices that help users—especially those in underserved regions—effectively harness geospatial data to drive actionable insights. Radiant Earth focuses on simplifying the process of integrating complex geospatial data into applications that address real-world problems, such as disaster response, climate change, and sustainable development, with an emphasis on making geospatial technology more accessible, intuitive, and impactful for a diverse, global user base.

Conclusion

The geospatial community is poised for transformative innovation. Our conversation with Jed Sundwall offered a compelling glimpse into the evolving geospatial landscape, emphasizing the urgent need for seamless data accessibility.

Developers and solution providers who have access to high-quality, interoperable data can easily integrate that into innovative tools, ensuring that the solutions created are more tailored, efficient, and scalable. This enhanced accessibility empowers not only large corporations but also smaller, niche providers who can leverage these standardized frameworks to address specific, localized needs. As a result, we’re seeing the democratization of geospatial technology, where anyone with the right data and vision can contribute to solving the world’s most pressing challenges.

By embracing diverse, localized solutions and fostering collaboration, we can pave the way for technologies that not only transcend maps but also integrate geospatial data into decision-making at every level. The future of geospatial accessibility lies in empowering communities and organizations to turn data into actionable insights, making the next generation of map engines not just possible, but inevitable.

We’d like to thank Jed for taking the time to share his viewpoints on the matter. We are confident that these insights and opinions, when heard and acted upon by other thought leaders, can create a ripple effect of growth in our industries. Be sure to watch the full podcast here.

Until next time!