Simulation models are typically used for making strategic decisions and our website includes dozens of case studies where we are able to see this strategic use of simulation. You can learn more about this from the Disruptive Business Strategies white paper.
Nevertheless, the current trend is digitalization, and thus people build digital twins – learn more from the Digital Twin Development white paper. Such digital twins are typically built to be used in the tactical and operational activities of a company. Therefore, a connection to an enterprise resource planning system like SAP becomes a critical task in any project like this.
Many clients of AnyLogic organize complex dataflow between their systems involving simulation models. This can be exported as stand-alone applications or used via AnyLogic Cloud API from a private Cloud installation on the company servers.
The operational simulation
Let’s dive deeper into this topic by looking at the steel manufacturing case study, that successfully went down this path.
For the goal of automation, the customer decided to develop a tool that could generate an optimized production schedule for each of the machines in the factory and forecast the systems behavior.
The customer required a production schedule and forecast with a horizon of at least one month. The objective for optimization was to maximize service level and fulfillment of the delivery schedule while considering resource efficiency. Key points of consideration include buffer capacities, transportation means availability, machine calendar, machine productivity and efficiency, and raw material availability.
AnyLogic flexibility allowed the client to customize the software tool precisely to the production management needs. What exactly does this look like?
The front end of the software was a customized solution, while the back end doing the heavy lifting was based on the exported AnyLogic model coupled with a tailor-made optimization engine. From a technological standpoint:
- the front end was based on JavaFX.
- the optimization engine uses methods such as local search strategies and Tabu search metaheuristic.
- the simulation model was implemented using an agent-based approach.
The simulation model and optimization engine organized a looping workflow that enabled an optimal schedule as well as a forecast for multiple metrics.
On the upper level, the solution was integrated in the production planning process based on the SAP ME system.
Output data from the simulation runs is given with intuitive charts, graphs, and figures, which can be exported to Excel files. From this, the optimal route for ordering material, production scheduling and storage based on demand forecast could be easily realized. Furthermore, demand forecasting from the simulations also provide insight on an ideal balance of manufactured goods.
We have highlighted this case study because it shines a light on the architecture of an operationally used simulation model. So, what we have seen and learned:
- The interface of the model can be done as a separate application (desktop or web-based), that may embed AnyLogic model animation, or plots from the AnyLogic Cloud.
- The model can operate in conjunction with an optimization engine/heuristic.
- The data is given from an enterprise resource planning system, where SAP is most frequently met.
- The typical goal of such usage is a kind of scheduling or schedule verification.
Subscribe to our monthly newsletter so we can bring our latest blog posts directly to your inbox.