Nokia develops network-aware route planning system for self-driving cars with Snowflake

Nokia is developing technology that could address one of the critical challenges facing autonomous vehicles: maintaining reliable network connectivity while navigating complex routes.

The Finnish telecommunications company is testing an AI-powered system that predicts optimal paths for self-driving cars, drones, and robotaxis based on network quality data. The initiative, still in pilot phase, represents Nokia's attempt to monetize its network infrastructure by making connectivity information accessible to autonomous vehicle operators.

Mikko Jarva, Head of Portfolio and Architecture at Nokia, says Snowflake’s technology is helping the telecoms giant predict optimal route paths for autonomous vehicles. These vehicles must maintain strong connections because network quality can affect a vehicle’s ability to navigate through traffic safely.

Jarva and his team work for a new unit at Nokia called Network Monetization Platform. The business unit has developed Network APIs, which bring mobile network capabilities within the reach of modern applications in pioneering areas of transportation, such as self-driving cars, robotaxis, and drone flights:

The Network APIs communicate through a platform with the mobile networks. We use the Network APIs to ensure that the route always has the best possible connectivity, and the mission can be safely executed.

Jarva and his team have combined the Snowflake AI Data Cloud with Nokia's Network as Code platform to create an optimal route path algorithm that predicts relevant network KPIs, ensuring autonomous vehicles have high-quality connectivity. 

Nokia’s relationship with Snowflake is based on the company’s acquisition last year of Rapid, the world’s largest API marketplace. When Nokia purchased Rapid, it also inherited its technology and talent assets, including a long-standing relationship with Snowflake. 

Nokia’s autonomous transportation initiative builds on the existing partnership with Snowflake. Jarva’s team stores information about API transactions across its Network as Code platform and enriches this knowledge with external data sources, such as footfall, traffic, and points of interest. The result is a data-enabled model for optimal route planning, he explains: 

The Snowflake platform has become a digital twin of the networking environment. And that technology can be used to provide optimized routes that maximize connectivity and ensure successful applications in autonomous transport.

Changing the network

The key to effective optimal route planning is the ability to combine data from multiple sources. Alongside data from networks, his team draws on information from publicly available sources and real-time data submitted to the platform by companies operating self-driving cars, robo-taxis, and drones, says Jarva:

All that information can be used through some clever analysis to predict the quality of the paths and provide better outcomes.

Devices, whether phones or self-driving cars, usually have no awareness of the network when they use its capabilities. Nokia’s project aims to change how devices and applications think about the network, he adds:

We want to make the network consumable and available easily through APIs. It's a niche area. I don't think many people know about it, but we’re ahead of the game at Nokia.

Nokia's initiative reflects broader efforts by telecommunications companies to extract additional revenue from their network investments. As traditional voice and messaging revenues decline, carriers are seeking new ways to monetize their infrastructure through data services and APIs.

When it comes to applications, Jarva gives the example of customers who use drones as first responders. These drones are connected to mobile networks to stream video for services, such as police operations. While these organizations have specific missions, they must comply with rules and regulations. 

Other potential applications include autonomous delivery and entertainment services, such as for consumers who require persistent high-level connectivity during a journey. Operating companies will hook into Nokia’s services through APIs. 

Jarva also cites the example of a robotaxi firm. These autonomous taxis generally know where they need to go and have several options for navigation routes. Nokia’s technology will help operators validate the routes through an API, showing which one provides the best pathway and uninterrupted connectivity. 

In another example, Jarva refers to network congestion, which can become a big issue for self-driving cars and robotaxis. Vehicles can get stuck in the middle of the road and create a traffic jam when they can’t communicate with remote supporting agents. Nokia aims to mitigate these concerns by allowing operators to call its services through integrated APIs.

While the potential applications of the technology are clear, it’s crucial to recognize that the pioneering services Jarva talks about are often still at a conceptual stage:

We’re at a pilot phase in terms of the capabilities. The wider use of Network APIs is starting to emerge. They can be used for other purposes, such as creating a silent second-factor authentication for more secure customer journeys. That use case is starting to happen in financial applications.

Advancing into new areas

Jarva recalls that the cultural and technological challenges of working with Snowflake were reduced because of the tech company’s long-standing relationship with Rapid. He says Nokia continues to explore new ways to work with Snowflake, including the development of Rapid Hub, an API marketplace that uses Snowflake technology.

The work of Jarva’s Network Monetization Platform business unit is part of a wider desire within Nokia to connect networks in what he refers to as a digital ecosystem. The aim is for clients to use Nokia’s services through this ecosystem. Some of those services will be supported by Jarva’s team’s work on route-planning optimization for autonomous vehicles:

The robotaxis and drone operators are already real use cases, but they are still a bit nascent. We believe the deeper integrations and first commercial applications will be in the prevention and identity verification space. Then, I think that making robotaxis and drone operations better by using our Network APIs and capabilities will happen in the next couple of years.

When it comes to lessons for other business and digital leaders who plan to use Snowflake technology, Jarva says value comes from breaking down data silos and using information from multiple domains in conjunction to create more value. He expects further developments during the next 24 months, including drawing on some of the innovations showcased at Snowflake Summit 2025:

Snowflake is great for breaking the data silos and managing access to data in a compliant way. And since we have our operational and business data for APIs in Snowflake, we are also looking into the more advanced capabilities of the platform to analyze performance, run future business scenarios, and figure out what we can do next.