Selecting the Optimal Machine Learning Framework for LiDAR-Based Railroad Inspection and Maintenance at Strukton Rail: A Comprehensive Evaluatio

The surge in machine learning (ML) applications has made robust ML operational frameworks imperative. Strukton Rail, a railway solutions company, faces significant challenges, including managing extensive LiDAR data, deploying models efficiently on the cloud, maintaining version control, and ensuring user-friendly solutions. In response to these challenges, this research not only evaluates but also provides a structured approach for selecting ML frameworks based on criteria relevant to the needs of Strukton Rail. This contributes to the domain by providing a blueprint for future ML framework selection in similar contexts. The study bridges the gap between ML model experimentation and real-world application, aiming to boost the efficiency of ML applications in practice. Overall, the framework deemed most suitable for Strukton Rail through this process was ClearML.