Localyte: Making Vehicles Smarter with AI

Italy

TL;DR:

  • Won 3rd in EIT Jumpstarter – Urban Mobility category
  • Improves vehicle sensor accuracy in real time
  • Completed MVP and launched PoC with partners
  • Targets OEMs and Tier 1 suppliers in automotive

Turning a Start-up Dream into a Driving Force for Safer Roads

What if vehicles could understand the world more clearly? That’s the question that sparked the journey of Localyte, an Italian deep-tech start-up founded with the ambition to build, learn, and solve.

From the start, the team knew they wanted to create a company, and then came the idea: improving the reliability of sensor data in vehicles. In a world where autonomous and assisted driving systems depend on precision, Localyte saw a clear gap and an opportunity to make a lasting impact.

How EIT Jumpstarter Accelerated Their Journey

In 2024, Localyte (then called Wyvernn) joined the EIT Jumpstarter programme in the Urban Mobility category. The experience was pivotal. “The programme helped us accelerate the commercial aspects of our development,” shares co-founder Valerio Lomanno. Through hands-on mentorship and tailored business training, the team refined their business model and built a network of industry insiders and fellow innovators. Taking home the 3rd place prize not only validated their work but also brought visibility within Europe’s mobility ecosystem.

The Technology Behind the Start-up

At its core, Localyte is a software-only solution that uses AI to improve vehicle sensor data—no hardware upgrades needed.

It tackles common issues like bias, drift, and scaling errors in sensors such as IMUs and odometers. Using a dual AI architecture that includes Error Filtering Neural Networks and Physics-Informed Neural Networks, the software dynamically corrects errors and fuses multiple sensor streams in real time. This makes navigation and ADAS functions more accurate and reliable, particularly in dense cities, tunnels, or areas with poor GNSS signals.

From MVP to Proof of Concept

During the EIT Jumpstarter programme, Localyte validated its Minimum Viable Product (MVP) in both real and simulated conditions. The results were promising: significant improvements in localisation accuracy. Their momentum continued with selection as finalists in the ESA–ANAS Grand Challenge—a strong signal of their technology’s relevance and potential.

The team is now running Proof of Concept projects with industrial partners, working toward commercial readiness through licensing models.

What’s Next?

Looking ahead, Localyte plans to finalise its PoC phase, secure pre-seed investment, and scale up through strategic partnerships with OEMs and Tier 1 suppliers.

Their solution not only improves safety and precision in modern vehicles but also supports sustainability by retrofitting existing hardware with software—extending the lifecycle of vehicle components without costly replacements. The team continues to push forward with R&D focused on next-generation sensor fusion models.

“We knew we wanted to build a company. The idea came later—but the vision to create something meaningful together has always been at the centre.”

Localyte is a clear example of how the right support at the right moment can transform an idea into a start-up with global potential.

Learn more about the EIT Jumpstarter programme.

More about Localyte: