The simulation platform to train and test AI for business applications

With AnyLogic simulation modeling you can construct detailed and robust virtual environments for training and testing your AI. The unique multi-method simulation capabilities provide a comprehensive tool for use in machine learning. Established in use at leading companies across industries, this fully cloud enabled platform with open API is enhancing and accelerating AI development today. Get the advantage in your development too!

Start now

Today, AI developers are working on an increasingly wide range of business applications in areas including manufacturing, logistics, supply chains, urban transportation, business processes, healthcare, asset management, and more. Successfully applying AI in these fields requires powerful and realistic virtual environments that enable risk-free training and testing of learning agents.

These environments need to be much more sophisticated than collections of images or videos, computer games, or even historical business data. They should also be able to simulate a rich spectrum of system behaviors, including those that have never occurred before, and interact with learning agents during the simulation.

Why use simulation for AI?

AnyLogic is the market and technology leader in the realm of general-purpose simulation. AnyLogic models help AI practitioners in many different practical applications, including:

  • 01

    generate synthetic data

    Generating unlimited amounts of relevant, clean, structured, and labeled synthetic input data for use in supervised learning, enabling machine learning, analytics and data mining applications.
  • 02

    train neural networks

    Training learning agents using deep reinforcement learning architectures in simulations of real-world scenarios. Simulation generates scenarios, allowing situations not captured or experienced in the real world to be trained on.
  • 03

    test trained networks

    Testing trained neural networks in a powerful, realistic, and risk-free environment. The behavior of trained systems can be safely explored and analyzed in a simulation environment.

Why use AnyLogic for AI?

  • Open API and programming capability

    The AnyLogic API is completely open to modelers and programmers. Support for across-the-board object-oriented development is fully integrated alongside AnyLogic’s visual modeling capabilities, giving developers the freedom to choose the most appropriate methods. Furthermore, simulations can fully self-configure using connected data.
  • AnyLogic Cloud

    The AnyLogic Company is the only simulation vendor with a fully functional Cloud platform. AnyLogic Cloud technology can help with model deployment by providing parallel model execution, easy sharing, and an open RESTful API for abstracting away from the implementation language. AnyLogic offers cloud access via subscription or private installation, with a choice of platforms either from cloud service providers or in-house.
  • Agent-based modeling

    AnyLogic’s unique agent-based technologies naturally position it as a powerful model building platform for training environments used in multi-agent reinforcement learning schemes. With the freedom AnyLogic gives to the creation and design of learning agents, the full range of cooperative, competitive, or hierarchical behaviors can be modeled.
  • State-of-the-art simulation technologies

    AnyLogic is the only platform that supports all simulation methods (Discrete Event, Agent-Based, System Dynamics) and any combination of them within one comprehensive development environment. Industry-specific libraries enhance and assist modeling for material handling, process modeling, fluid, rail, pedestrian, and road traffic, all working seamlessly with each other. GIS maps integration connects simulation logic with geographical locations, roads, routes, regions, and more.
  • Ease-of-use

    AnyLogic allows the use of visual modeling languages, scripting, and extensibility. Natively supporting flowcharts, state-charts, action charts, and stock and flow diagrams, eases the development process, while native support for scripting maintains accessibility. Complexity can be managed, and the model building process kept simple, clear, and elegant. AnyLogic easily connects with DL4J, RL4J, SKIL and more.
  • Proven application in a variety of industries

    As a result of the comprehensiveness and customizability of AnyLogic, it is the de-facto simulation platform for Fortune 500 companies across the globe. The familiarity of these businesses with AnyLogic is advantageous to AI practitioners who want to use the platform as part of AI solutions for these companies. If you are interested, check out the customer presentations from our annual conference and find out about their experiences and use cases.

Andrei Borshchev, CEO, The AnyLogic Company, at the 2019 AnyLogic Conference, summarizing the simulation application space for machine learning. Highlighting the risk-free and ease-of-use for training that AnyLogic provides.

For a much deeper dive into an example, you can watch the Arash Mahdavi and Ty Wang presentation.

Don’t take our word for it, find out for yourself!

Fill out the form and you can access a range of materials, including an example reinforcement learning system with source files and instructions, so that you can get a clear idea of how simulation modeling with AnyLogic can benefit your AI development. See the videos, hear the discussions, read the reports, and get some hands-on!

Materials include:

  • Example RL model with source files and documentation.
  • PwC presentation videos and paper
  • Agent based simulation + Reinforcement Learning: AnyLogic Conference video

Webinar: Machine Learning Using Simulation

Join us as we explore the benefits of training a machine control policy using
deep reinforcement learning in a simulation modeling environment.

July 11th, 15:00 UTC. Find out more…