Finnish Transport Infrastructure Agency has to manage data generated by over €20B worth of infrastructure assets—and find a way to share it with other agencies and contractors. Here’s how it simplifies sharing, streamlines reporting, and cuts costs with Snowflake’s platform.

Väylävirasto , Finnish Transport Infrastructure Agency, is responsible for €20B worth of infrastructure assets, including 6,000km of railway lines, 78,000km of highways, 6,000km of pedestrian and bicycle lanes, 8,300km of coastal fairways, and 8,000km of inland fairways.

Managing and maintaining such a broad, complex infrastructure portfolio creates and consumes large volumes of data. And it also demands flexible collaboration with other agencies, including Fintraffic and Traficom, two bodies responsible for traffic control and regulation.

So, how did FTIA scale up its data infrastructure to support these large data volumes—and the need for seamless collaboration?

FTIA continues its transformation journey in the cloud

FTIA has always embraced change in its efforts to improve operations. “When I first joined, three agencies (one for waterways, one for rail, and another for roads) had just been combined into one,” explained Tomi Mykkänen, Enterprise Architect at FTIA . “So, we had three different on-premises systems based on Cognos and Oracle.”

Merging these on-premises systems was the first of many infrastructure projects Mykkänen has helmed for FTIA . He also spearheaded the deployment of on-premises Hadoop architecture in 2014.

But in 2018, the agency began a bigger strategic move to a cloud-based architecture. Solita, FTIA’s longtime technology partner, delivered a project incorporating Snowflake and Tableau.

Once Mykkänen saw the project running, he knew that those systems would offer a solid foundation for FTIA’s cloud data platform. “We looked at many different BI approaches, and I’d worked with some solutions in the cloud before,” explained Mykkänen. “But once I saw Snowflake and Tableau running, I decided that was the way we should go.”

Strong first impressions lead to strong results

Even before full deployment, Mykkänen was pleased with how easy it was to migrate to Snowflake. “It was much easier to migrate to the cloud with new tools like Snowflake,” said Mykkänen. “And it was far simpler than it would’ve been to keep upgrading versions of on-premises software.”

Today, FTIA now ingests data through Amazon S3 before orchestrating that information through Solita’s custom agile data engine. That makes the data mart models FTIA uses in Snowflake’s platform. From there, the agency relies on Tableau and IBM Cognos to visualize and analyze its data.

This single data platform offers a potent mix of trusted insight and cost savings. “The biggest benefit is that the data is all in one place, and it’s easy to connect that data through different tools,” explained Mykkänen. “It also saves us a lot of money. Storage is so cheap in Snowflake and AWS, which helps us free up budget we can use elsewhere.”

FTIA ingests all kinds of infrastructure and vehicle data, including information on weather conditions, infrastructure maintenance, asset conditions, and geolocation data for trains and vessels. Combined, these datasets allow deeper reporting across FTIA, other government agencies, and their contractors. “There’s day-to-day operational decision-making going on, like what conditions the roads are in, for example,” said Mykkänen. “But we also indirectly employ around 14,000 people across all projects. This data is also vital for giving them the information they need to properly understand our requests and maintain our infrastructure.”

Data sharing enhances collaboration across agencies and contractors

Sharing data with organizations has quickly become a key part of FTIA’s data operations.

For example, FTIA sends road maintenance information to Fintraffic to help it manage traffic control, and it receives valuable data back that it uses to better manage maintenance work—all through Snowflake’s data sharing capabilities.

The agency has also started sharing maritime data with Trafficom, and Mykkänen was thrilled at how quick it was to set the agency up with a reader account in Snowflake. “They aren’t even a Snowflake customer themselves, but the process was so quick,” explained Mykkänen. “It took around nine minutes to set everything up—and I think it could be done even faster than that.”

And now it’s up and running, Snowflake’s data sharing infrastructure allows for faster data processing with datasets that are trusted. “It’s much faster than ETL/ELT, and easier,” said Mykkänen. “Best of all, we’re all working from the same dataset. It’s one source of truth, and it means we don’t have to constantly model the data ourselves, so it takes away a lot of workloads.”

The speed and simplicity of data sharing between FTIA and Fintraffic and Trafficom has triggered a cultural shift across other agencies. Now, other transportation management agencies in Finland are also considering data sharing to orchestrate railways and air travel and deliver end-to-end transportation management.

Looking ahead with Snowflake’s Data Cloud

“We’ve been working with Snowflake for a few years now and have recently extended our contract as part of our long-term partnership, so we definitely have plans for the platform,” Mykkänen revealed.

Having shown how FTIA is using its platform at a recent Snowflake user group, other government agencies have approached Mykkänen about data sharing. 

The agency is also looking to integrate more data with Snowflake’s platform as part of its continued push for centralized data governance. A central, scalable platform also fits in with a broader strategy to create digital twins of Finland’s road, rail, and waterway infrastructure.

“Our digital twin strategy could help us improve the way we handle all sorts of processes,” Mykkänen explained. “Predictive maintenance is a big area where this could help, but it will also allow us to prioritize work better and align maintenance projects with our overall strategy. It all revolves around Snowflake and the data we store there.”