Continuous Development and Safety AssurancePipeline for ML-based Systems in the Railway Domain

This paper details the implementation of a Machine Learning Operations (MLOps) process for Automated Driving Systems (ADS) in the railway domain. It addresses the challenges of applying machine learning in safety-critical railway systems and outlines a comprehensive approach to continuous development and safety assurance of ML-based systems. The paper emphasizes the use of Git-centric methods and appropriate tooling to automate the process and ensure safety.  

Download the full whitepaper to understand how to implement a safe MLOps process, including data quality assurance, ML model development, and safety case management, ensuring the trustworthiness of AI-based functions in driverless trains.