In manufacturing, time is never just time; it is capacity, cost and competitive advantage compressed into a single variable. When demand spikes, orders pile up and due dates loom, production managers face a deceptively simple question: do you train your workers to cover multiple stations and move them around, or do you simply ask them to stay later? The answer, it turns out, is far from obvious.

Overtime is one of the most widely used levers in production management, yet it remains one of the least studied. It can slash manufacturing lead times dramatically, but applied wrongly, it creates a new problem: some workers standing idle while others are clocking extra hours, burning budget without burning through the backlog. Worker cross-training and flexible reallocation offer an alternative, but come with their own costs and constraints. The real question is not which tool is better in theory: it is how to decide between them in practice and when to combine them.

This thesis tackles that decision head-on. Starting from a structured literature review, the research builds a formal model of capacity adjustment in manufacturing flow shops, then stress-tests it through Discrete Event Simulation and Python-based analysis. The output is not just an academic contribution to a surprisingly underexplored corner of operations research; it is a decision framework grounded in how real production systems actually behave under pressure.

If you are drawn to the kind of research where mathematical rigor meets shop-floor reality and where a well-built simulation can reveal insights that years of intuition missed, this thesis was built for you.