Measuring to decide: job order costing and assembly redesign in an engineer-to-order company.

In Engineer-to-Order (ETO) production, each job differs, making labor hours hard to trace from aggregate records. This thesis examines how recording direct and indirect workshop hours by job and phase improves visibility of labor use and supports operational decisions. It studies Alpha, an anonymized Italian manufacturer of end-of-line machinery, as a single embedded case. Sources include hour records, phase master data, plans and work sampling in conveyor assembly (624 observations in 40 rounds over 21 days). The previous attendance-based system reconciled paid time but did not reliably identify the job and phase worked or separate direct from indirect time. The new system assigns attributable hours to jobs and phases; phase targets support job Gantt charts and a view of planned load against capacity. A fully configured reference machine has up to 102 possible phases and 2,989 target hours; each order uses only the phases it requires. In conveyor assembly, the share of observations classified as working was 68.4% at a fixed workstation and 45.5% in line assembly. Waiting, transporting and walking were not observed at the workstation, while material handling was more frequent. These descriptive results supported the workstation pilot and point to material feeding as the next issue to examine. The thesis reports labor hours and activity shares; it does not calculate the full monetary job cost or establish a general causal effect.