Synthetic data-based OEE evaluation in a matcha manufacturing SME

Abstract

Small and medium-sized enterprises (SMEs) in food manufacturing often lack systems such as Enterprise Resource Planning (ERP) or Manufacturing Execution Systems (MES), which makes it difficult to measure production performance. This thesis presents a simple way to evaluate Overall Equipment Effectiveness (OEE) for a matcha powder production line using synthetic data instead of real factory data.

A hypothetical small matcha production line is defined based on literature. The key variables for OEE calculation are specified, including planned production time, downtime, cycle times, total output, and defects. Using Python and Excel, a 30-day synthetic dataset is generated with rule-based logic to mimic realistic behaviour. OEE is then calculated for each day through its components: Availability, Performance, and Quality.

The results suggest that, in this simulated case, cleaning time, minor stops, and reduced speed operation have a strong effect on OEE, while quality losses are smaller but linked to moisture and temperature conditions. The numbers are illustrative, not real factory values, but the approach shows how SMEs can use synthetic data and basic tools to learn OEE and prepare for future digitalisation.