Pain-free data integration with Overture Maps, GERS & Databricks
.png)
You’d think integrating spatial data would be simple, wouldn’t you? Two features coincide spatially, so you run some form of join on them and continue with your workflow using your perfectly integrated data. The reality? Not so perfect.
Geospatial data is often difficult to integrate - particularly at scale. Even when licensing is straightforward, onboarding and standardizing across datasets can take weeks - slowing down time to value.
Overture Maps Foundation has recently made a HUGE addition to their global open data products to eliminate this pain - the Global Entity Reference System (GERS). GERS provides consistent identifiers for geospatial features like roads, places, and buildings for frictionless data integration. In this post, we’ll do a deep-dive into how you can use GERS in your Databricks environment with CARTO to perform repeatable spatial analysis without the usual overhead.
Overture Maps is an open data project from the Linux Foundation that publishes validated, production-ready geospatial datasets. Maintained by members like AWS, Microsoft, Meta and TomTom, Overture Maps data includes core geographic features - roads, buildings, places, and boundaries - at a global scale, released monthly. The data is structured for interoperability, with consistent schemas and changelogs.
We’ve been working with Overture Maps to make this fantastic data available on cloud platforms like the Databricks Marketplace, to remove painful download and ETL processes. Keep reading to learn how you can access Overture Maps data, the cloud-native way!
GERS solves a long-standing challenge in geospatial analytics: aligning features across datasets that use different schemas or geometries. Instead of relying on heavy spatial joins or custom conflation logic, users can perform simple column-based joins using GERS IDs. The key benefits of this approach are:
- Open and system-agnostic: GERS IDs are freely available and not tied to a specific vendor or API.
- Globally unique: Every ID is a UUID, ensuring it can be referenced across systems without collision.
- Stable and versioned: IDs persist across data releases, making them safe for long-term use in analytics pipelines.
- Interoperable: Bridge files are available to match GERS IDs with identifiers from sources like OpenStreetMap and Meta.
Overture data can be accessed directly within the Databricks environment via Delta Sharing. This allows users to work with up-to-date geospatial datasets - such as the Overture Roads layer - without needing to download or manage files manually.
At CARTO, we’ve made this data available on the Databricks Marketplace, so you can easily build it into your analysis.

By leveraging CARTO with Databricks, you can analyze Overture Maps data through a fully cloud-native pipeline where spatial analysis is performed directly on shared, standardized geospatial data - no ETL, exports, or third-party file formats required. Sounds good? Let’s check it out in action!
Want to start using Overture data for more scalable and consistent analytics? We’re going to be walking through how you can do just that, using the example of analyzing delivery delays across a set of warehouse-to-store routes. When a delivery misses its delivery time - or Service Level Agreement (SLA) - the “why” isn’t always obvious. Was it traffic? Route choice? Dwell time? Without standardized spatial data, solving that costly problem at scale is painful and slow.
This tutorial walks you through how to identify the causes of delivery delays using CARTO, Databricks, and Overture Maps data.
Follow along by signing up for a free 14-day CARTO trial here!
Step 1: accessing Overture data, and your own
You can set up a connection to your Databricks lakehouse in CARTO to access any first-party data. You can do this easily in the Connections tab of your CARTO Workspace (see below).

You’ll also need to subscribe to the Overture data that you need - for this example, these are the destinations for our routes; pharmacies within 1 hour of our Baltimore warehouses. In the Databricks Marketplace, search for Overture Maps - Places, select Get Access, and follow the instructions to subscribe to the data in a location of your choice.
Step 2: Establishing delivery routes with CARTO
Now, you’ll need to establish the delivery routes between the warehouse and stores. The easiest way to do this is by executing the below analysis in CARTO Workflows, our low-code tool for automating complex spatial analysis which executes entirely in your lakehouse.

The workflow above creates a routing matrix: on the top left it imports and filters the Overture Places table to a list of destination stores, and on the bottom left it imports a list of warehouse origins. The Cross Join step makes a full set of warehouse/store origin-destination pairs, then passes it to the Create Routes function to identify the optimal route between all origin-destination pairs, which are then saved and visualized on the map below.

The results of this analysis are structured outputs from CARTO’s Location Data Services (LDS) routing engine: a list of warehouse/store routes following roads and highways. How can we analyze these routes to identify the impact of traffic incident points on deliveries? Enter GERS!
Step 3: GERSification!
With GERS, each segment of the road network (from the table Overture Maps - Transportation) is assigned a unique ID. By matching both our LDS route outputs and our traffic incident points to the GERSID road segments beforehand, we now have a universal reference framework. Every vehicle ping, delivery, or delay can be traced back to a precise, shared segment.
Step 4: Identify traffic-related failures
Once data is joined by standardized road segments, we can isolate late deliveries that correlate with traffic incidents. Interactive widgets in CARTO Builder allow quick filtering and comparison between late deliveries with and without known traffic events (see below).

Step 5: Turning analysis into action
With traffic-related issues clearly identified, you’re left with a high-value subset of SLA misses that demand a closer look. These are the deliveries that should have arrived on time based on route and road conditions, but didn’t. Rather than chasing false leads, teams can focus on these unexplained issues - and that’s often where the biggest performance gains are hiding.
Geospatial analytics shouldn’t be slow, siloed, or speculative. With GERSified Overture Maps data, CARTO’s cloud-native platform, and Databricks as your lakehouse, you can shift from manual, one-off investigations to a repeatable and scalable process for uncovering delivery issues.
We're looking at how we can enhance and advance our support for Overture Maps data across the CARTO platform, and we're looking for partners to collaborate on this with for input, testing and feedback. If you have an interesting use for Overture Maps data and would like to help shape the future of how this is used, please get in touch!