Fatigue Assessment for Transportation Engineer Determination

Fatigue among locomotive engineers poses a persistent safety risk within the transportation sector, exacerbated by long shifts, monotonous environments, and high cognitive demands. To address the limitations of current fatigue management practices—largely reliant on subjective self-reports and periodic evaluations—Mosaic ATM proposes the development of a Human Factors Fatigue Assessment System. At its core is the Fatigue Assessment for Transportation Engineer Determination (FATED) tool, a multi-feature, machine learning-driven system designed to detect real-time behavioral indicators of fatigue using non-invasive video data. By analyzing factors such as eye closure, gaze deviation, head posture, facial micro-expressions, and body movement, FATED generates a continuous, objective fatigue score. The system architecture features a lightweight, multi-tiered model optimized for edge computing hardware, ensuring real-time operation within locomotive environments. Phase I will focus on establishing the technical feasibility of this approach through algorithm development, validation against open-source datasets, and testing in automotive contexts as a precursor to rail applications. The project builds on prior work in driver attention monitoring and leverages human factors expertise to create a scalable, transferable framework for improving transportation safety. HFFAS aims to reduce fatigue-related incidents and enhance situational awareness across the rail industry and beyond.