Kognic Mobilix Partnership to Advance Physical AI Data Flywheels

Kognic and Mobilix Partner on Physical AI Data Flywheels

On August 10, Kognic announced a partnership with Mobilix, a Korean company specializing in autonomous driving and mobility, to enable the construction of autonomous driving and physical AI data flywheels. The collaboration will provide customers in South Korea with data flywheel capabilities designed to improve how training and validation data is extracted, triaged, and processed. The data is collected from autonomous vehicles and robots operating in real-world environments, where increasing data complexity and volume create significant challenges for development teams.

Data-Driven Improvements for Autonomous Driving Operations

The partnership will make data-driven functional improvements visible to teams responsible for Mobilix’s Robo-bus and Robo-taxi operations, as well as developers working on Level 2+ ADAS and autonomous driving systems. The companies intend to build iterative structures around actual operational data, allowing development teams to address quality requirements at scale while improving their understanding of how data relates to model behavior. The approach is also intended to support the identification of unexpected situations, unusual behaviors, and rare events that can be difficult to capture through conventional development and validation processes.

Supporting Higher-Degree Automated Driving Deployment

By combining operational data with structured processes for extracting, triaging, and processing training and validation information, the collaboration is aimed at addressing the growing demands associated with more automated driving systems. Both companies plan to rapidly engage with customers committed to deploying or operating higher-degree automated driving technologies. The partnership therefore focuses not only on handling larger quantities of real-world data, but also on creating iterative feedback structures that can help teams identify critical scenarios, evaluate system behavior, and use operational insights to support ongoing improvements in autonomous driving and physical AI applications.

Frequently Asked Questions

What is the purpose of the Kognic and Mobilix partnership?
The partnership is designed to help customers build physical AI data flywheels for autonomous driving and robotics applications using real-world operational data. These structures are intended to improve the extraction, triage, and processing of training and validation datasets while helping development teams identify rare events, unexpected behaviors, and complex relationships between data and model performance. The collaboration will support Mobilix’s Robo-bus and Robo-taxi operations, along with Level 2+ ADAS and autonomous driving developers seeking to deploy higher-degree automated driving systems.