In 2017, we released AnyLogic Cloud to the public. With a userbase of around 10,000 modelers, it is now the largest public online platform for people working in simulation.
The public version of the service is available to everyone, allowing them to run models and experiments. Online models and experiment results can be shared online and run from a browser. The platform leverages the advantages of cloud computing: this way complex experiments are performed faster than on a regular computer, and the quality of animation depends minimally on the power of your device.
For business tasks, AnyLogic Cloud is available by subscription or installed in a private cloud, on a company's infrastructure (this version is called Private Cloud). These versions expand the capabilities of the service: no time restrictions, third-party connectivity, Cloud API access, and security configurability to ensure service in accordance with corporate security requirements.
In this article, we have collected 5 typical scenarios for using AnyLogic Cloud from our Private Cloud clients. These scenarios tell how clients implemented the service and what tasks they solve using it.
The cloud for storing models
When a company has several people working on a model at the same time - for example, a simulation modeler, a project manager, and a department director - it is convenient for everyone to work from a common repository. So, the modeler can change the model and instantly deliver the changes to the customer, and the manager and the director can test ideas and evaluate them without installing AnyLogic on their computers.
One example case is the planning and optimization of enterprise processes using simulation modeling. Developers create a process model and upload it to the Cloud so that enterprise managers can test hypotheses, set up experiments, and propose updates to the model. As soon as engineers have updated the model in the cloud, its new version is immediately available to managers. This process significantly reduces development time.
When a model is ready, a manager can share it, for example, with the production manager. You do not need to install AnyLogic for this: simply share the model’s link and it can be accessed from any device with a browser. Continuing the example, the production manager analyzes the results of experiments that have already been run and also launches new ones before uploading their results in the form of reports. The Cloud API allows you to connect the Cloud service to third-party platforms such as Power BI or Tableau and take advantage of their advanced possibilities for visualizing and analyzing simulation results.
Online model-driven solutions
Some companies are using cloud models to help develop custom digital tools that are specific to their industry. They connect cloud models to external data sources, create custom interfaces for them, and integrate with enterprise analytics platforms to visualize results. The advantage of such solutions is their versatility. Solutions can be embedded into business processes of any complexity and used to solve specialized tasks, and their interfaces can be extended and customized for users.
Here is a demo example of a solution for analyzing the operation of automated guided vehicles (AGV) in an enterprise.
Cloud for consulting
Simulation modeling consultants develop models for companies and integrate them into business processes. To maintain these models after implementation, consultants upload updates to the cloud, where the customer is given access to them: access levels and user rights can be managed at an account level. At the same time, the customer does not need to have an AnyLogic license: the cloud model simply launches in a browser using a link.
Consultants often develop online demo models to solve common problems in various industries. For example, scheduling and promoting an online store or optimizing daily tasks in a call center. When a customer is given access to an online model, they can analyze key indicators and rapidly develop changes for current processes. The remote developer-consultant, in turn, expands the model with additional functions and updates it in Cloud so that changes instantly become available to the customer.
Working with large models and complex experiments
Working with detailed simulations of complex systems usually requires large computational resources. If a company creates a resource-intensive model for day-to-day work planning and hypothesis testing, it could run out of computing time. One of the reasons for slow-running models and experiments can be that they are being developed and run on one computer: the machine's resources are too occupied by parallel tasks and the time needed to run a simulation increases.
To accelerate model development and more quickly obtain simulation results, some companies choose to run their simulation models in AnyLogic Cloud. Thanks to the nature of cloud computing, multi-run experiments in AnyLogic Cloud are faster and more efficient than on a regular computer because they are not limited to the resources of one machine. AnyLogic Cloud responds to the need for computing power by bringing additional nodes and cores online as needed.
Cloud helps companies with resource-intensive models:
- save on the purchase or lease of additional computational capacity and its installation;
- make more informed decisions more quickly;
- keep a database of all conducted experiments - with both input data and results.
AI training with cloud models
Companies that create systems based on artificial intelligence use models to experiment with them, to obtain synthetic data for training, and for testing AI algorithms. To get the results of experiments faster, they run them in the cloud, and for remote control and parallel execution they use AnyLogic Cloud’s RESTful API for Python, Java and JS.
Dr. Ali Azgari recently took advantage of AnyLogic Cloud to create a machine learning based mass vaccination simulation and application that is now in public use. To do this, Azgari built a simulation model in AnyLogic, uploaded it to AnyLogic Cloud and, using the API, conducted a series of experiments. Data from the experiments was used to train algorithms in an AI model. Then, for the convenience of users, Azgari created a web application based on the results of the research. See how it works.
AnyLogic Cloud is a powerful online tool for simulation models. Any AnyLogic user can upload their own model and run it in AnyLogic Cloud – try it! Learn more about its functionality in our blogs:
- What is AnyLogic Cloud? Service overview
- How to use the AnyLogic Cloud API in Python
- How to create web interfaces for models
If you have any questions about how to use the service for your business tasks, get in touch!