Processing incomplete data for simulation-based production planning in shipbuilding

Abstract

This paper presents an approach for simulation experiments at early stages of the shipbuilding process. In the early stages of the planning process, the design of the ship is not yet complete. However, an early scheduling of the production is already needed for staff employment and material ordering. Because the plan is based on incomplete input data, the uncertainty in production planning is much higher than in serial production with fully detailed product design and routings. The uncertainty reduces the accuracy of planning and, therefore, has a significant impact on its quality and robustness. The paper presents an approach to describe this uncertainty and to reduce it by generating design details from historical data.

Keywords:

• production planning
• incomplete data
• uncertainty modeling
• one-of-a-kind production
• throughput time estimation

Statement of contribution

The early planning phase in the one-of-a-kind production is characterized by a lack of construction data and therefore a high uncertainty with regard to processing times. The following paper describes an approach on how to cope with this uncertainty by on the one hand describing and modeling it, and on the other hand creating more transparency in processing time estimation and reducing the uncertainty in the processing times.

The new approach for modeling the uncertainty is taking the data quality into consideration to determine the possible range of values. It is divided into three steps: to identify uncertain planning data, to quantify uncertain planning data and to model the impact of uncertain planning data on the target values.

The new approach for creating more transparency in the processing time estimation is splitting the estimation process by first estimating a bill of material and then linking the bill of material to process times. Therefore tracking estimation errors gets easier, as well as the estimation gets precise. The bill of material also enables the early use of simulation in shipbuilding, using the Simulation Toolkit Shipbuilding (STS).