Often, people hear “custom” in the technology sector and, especially if they are relatively new at this game, they become nervous, imagining complicated scenarios, in-depth code writing, and hours spent trying to figure out left from right. But it doesn’t have to be like that, and this is illustrated in the webinar given by Arash Mahdavi and Tyler Wolfe-Adam.
- About Enterprise Simulation
- Why Custom Libraries
- What are Custom Libraries
- SIR Model Case
- Real-World Case Studies
- Conclusion and Video
What is enterprise simulation?
In this webinar, the presenters wanted to connect the concepts of model building at scale for enterprise simulation and custom libraries in AnyLogic. So, let’s first look at enterprise simulation. It has the following characteristics:
- The objective of enterprise simulation is the creation of numerous simulation-based digital twins representing the components of a distributed business, for example hundreds of branches of a fast-food restaurant spread throughout a country.
- Each component is a separate digital twin. The models representing these business units are different in details, but they all share common business applications. Each fast-food restaurant, for example, could have different menus or layouts, but they will all follow the same business model.
- All the models that are digital twins of each business are deployed on servers and are accessible by the appropriate decision maker. The level of access to each component would be managed and maintained by an administrator.
So, in a nutshell, enterprise simulation is a group or set of models with common underlying behavior that are accessible by a large number of stakeholders. The objective is to create many simulation-based digital twins.
Why custom libraries?
So, if you have a number of simulation models from a number of distributed businesses, it makes sense that these models would have some similarities, even though they can vary a little at each site. Therefore, a key factor of enterprise simulation is the reusing of common or generic elements. Why create something new every time when you can just reuse what already works – in other words, why reinvent the wheel.
Instead of creating a new model for each component, here you can use a custom library. This potentially complex library can be created by a few advanced modelers, and then other modelers can just alter behaviors.
What exactly are custom libraries and how to create them?
Custom libraries are simply an AnyLogic model that can be made available to all future models that you would like to build.
Creating a custom library is straightforward. Basically, add a library object to your existing model, check the agents that you want to include and then export this library. Within this, you can choose to create documentation automatically and even add the library to your palette. At the same time, you are also able to create an icon that will show up on the side in the library as other icons do.
The output library is a .jar file and this is what will be used to share your custom library. All someone will need to do is go to their AnyLogic palette and add a new library, choosing this relevant .jar file. Now, it can be used.
You can see an example of how this is done in the webinar, which you'll find at the end of this blog post.
Use case for types of custom model assets
In the webinar, you can find a few different examples of creating these custom libraries, but here we will just talk about one. An SIR model was created with a focus on one clinic with some parameters added and where some experiments could be created and analyzed. This is a very simple model that could do its job and then be done with.
But what if there was another clinic which had a similar focus? In this situation, the first model has already been built and can be used in the development of further models for this other clinic. In the webinar you will find more details about this example.
Real-world custom libraries
Many companies have already created their own custom libraries, either for their own idea of enterprise simulation or because it makes future creation of different models easier and more convenient because they have their own personalized and ready-to-go libraries.
In this case study a library was created for agent-based modeling with regards to active shooting events. This library shortened what a new researcher had to do when starting work on model development. It resulted in students being able to focus on research and not on the intricacies of the model.
When The Port of Long Beach, California had a plan to invest $1 billion dollars, they worked with TranSystems to quantify systems requirements to meet capacity goals. It was necessary to model the rail system of the port using AnyLogic’s Rail and Fluid libraries. In addition, TranSystems created their own custom libraries to further ease the modeling process.
Machine Centers Manufacturing and Engineering Group worked together to develop a flexible manufacturing systems (FMS) simulator for manufacturing optimization. To create this simulation, digital twin experts at Engineering Group made an FMS-specific simulation modeling library. This library was made up of a collection of reusable agents and Java classes related to FMS applications. This was used alongside the standard AnyLogic libraries so as to allow for rapid prototype facility designs.
Custom libraries and enterprise simulation are two interconnected ideas, as we have seen, and together they enable model development to be quicker, smoother, and more accurate. This ensures that businesses can operate effectively and efficiently without having to worry about the small details. Custom libraries have huge potential and when implemented correctly can pave the way to success.
Here is the webinar for you to watch in full.