PROJECT PROFILE

Model Assisted Production Planning

BlueScope is Australia’s largest steel producer. Its Port Kembla plant is a vertically integrated manufacturing facility incorporating iron making and downstream finished products including household brands such as COLORBOND® steel, strip steel coils and flat steel plate products.

BlueScope initiated a project aimed at improving performance of its flat steel product division. Adaptint was engaged to assist with the development of a digital twin (simulation model) of the plate manufacturing process including both the rolling mills and downstream plate finishing processes.

The digital twin is tightly integrated with a SQL database that is used to load real-time production schedules and inventory data into the model. The SQL database enables the user to configure plant operating parameters such as shift patterns, asset availability and user configurable materials handling sequences. Collectively, the Digital Twin and SQL database comprise a suite referred to as “MAPP” – Model Assisted Production Planning

MAPP is designed to assist production managers and supply chain personnel to test production runs before they are scheduled in the real plant. Real-time inventory and WIP can be uploaded to the Digital Twin enabling the model to reflect the actual conditions in the plant. The model then processes the rolling schedule and assesses the performance of the plant including surge capacity, resource utilisation, and key statistics such as:

MIFOT (Manufactured In Full On-Time)
AIFOT (Ready for dispatch in full on time)
DIFOT (Delivered In Full On Time)

A key feature of the model is its ability to simulate resource levels with the use of integrated shift planning. This enables production managers to run manpower planning scenarios that aid in understanding the impact of staffing on intermediate product stockpiles, WIP (work in progress) and delivery timelines.

The model is incorporated into everyday production planning. It enables supply chain managers to verify production schedules before they are manufactured and adjust a schedule if the model indicates the schedule may not meet performance expectations.

This use of a digital twin is a good example of the benefits of simulation tools in a modern manufacturing setting. Simulation models can be used to test and analyse the efficacy of a planned modification to a plant, to identify bottlenecks, assist in production planning or to provide insight on the anticipated benefits of a capital expansion project and more.

What's Unique About This Digital Twin?

The Plate Mill Digital Twin is an agent based, multi-method model utilising several AnyLogic simulation modelling libraries including the following:

Discrete Event Simulation
Material Handling Library
Rail Library
Road Traffic Library

A key feature of the model is the extensive use of overhead cranes to move plates throughout the processing and plate dispatch areas. In the real world crane operators choose where to move a plate based on its attributes; process phase; destination; dispatch point; length; gauge etc. The model endeavours to mimic the decision process of crane operators by reading the the plate's unique parameters then searching through exiting stacks for plates with matching attributes. Thus the destination node for each movement is calculated at the moment the crane picks up the plate agent. Controlling and coordinating the multiple crane bridges and overhead cranes throughout the facility was one of the challenging aspects of this project.

The model also makes extensive use of Anylogic's built-in database. Each plate is identified with a unique plate ID that is linked to a specific customer order. The model tracks and records the location and movement of each plate throughout the plant from the point of exiting the furnace doors to being loaded onto outbound transport. Twice a day the model generates a plate dispatch plan based on the finished product inventory and automatically schedules outbound logistic movements for both train rolling stock and trucks along with a loading list for each respective vehicle that includes the plate ID for each plate to be loaded. At loading the model compares the load out date with the planned delivery date (DXW) and calculates the plant's DIFOT statistic - a crucial supply chain KPI. It also produces a comprehensive dispatch and inventory report that can assist production managers to forecast stock levels and the impact on warehouse space and external plate storage capacity.

Could You Benefit From a Digital Twin?

Simulation is a powerful business tool that can help you to improve profitability, reduce costs and get more out of your capital investment. If you would like to learn how simulation modelling could assist you, please contact us.

Home