If you have not yet downloaded AXON Predict, make sure you do that before following this installation process.


  • AXON Predict downloaded to the machine on which you will install.
  • Docker (and Compose) are required. Make sure they’re already installed and configured for your environment.
  • You should be familiar with Docker technology and know how to use Docker commands.
  • System Requirements: 
    • Minimum requirements for Docker: version 18 of Docker and version 1.22 of Compose.
    • Minimum requirements for RAM: 6GB available RAM to run the AXON Predict containers. (When running Docker in Apple Mac or Microsoft Windows environments, allocate a minimum of 6GB of memory to the VM running Docker.)
    • Bash is required to use the script (.sh) provided in the download.


Instructions to Configure and Start AXON Predict

Configure AXON Predict

  1. Go to the local directory containing the AXON Predict download. You’ll see a single ZIP file named for the version you downloaded.

  2. Unpack the ZIP file. You’ll see files similar to the following:

  3. First, you need to edit The changes you make depend on the environment in which you will use AXON Predict: either as a simple, single server install or as a cluster install (with mesos/kubernetes).

    For this installation, we will configure for a simple, single server.

  4. Open In the "publicHost=" property, replace the default domain name shown ( with IP address for the machine on which you will deploy AXON Predict. DO NOT use the name "localhost", but instead you must specify an actual IP address. Leave all other properties at their default values and save the changes.

  5. Now, still from this directory, run the script In Windows, you can double-click the script name. (If using another operating system, you can open a terminal window and run the command: bash

    This adds your environment settings to the configuration templates provided in the download and creates the new directory, configs. The configs directory contains 3 subdirectories: composekubernetesmesos. The Compose directory contains the docker-compose files for AXON Predict.

    (Note: Although the docker image includes kubernetes and mesos, for the purposes of this simple installation we will ignore those subdirectories.)

  6. Change to the /configs/compose directory. To use the Quick Starts, you need to modify the atom-compose.yml to identify the CSV-Adapter (data-loader) that will provide the simulated data for the Quick Starts.

  7. Open and edit atom-compose.yml as follows:
    1. Locate the data-loader section and, within that, the "image" property.
    2. Replace the default value in this image property with:
      (You can copy and paste from here.)
    3. Save and close the file.

    IMPORTANT NOTE: The AXON Predict UI is defined, by default, to use port 80. If port 80 is not available to the UI, you need to modify cloud-compose.yml to specify an available port for the UI, as follows:

      1. Open cloud-compose.yml.
      2. Locate the ui section and, within that, the "ports" property, similar to the following:

            - "80:80"
      3. Replace "80:80" with an available port. For example, to change the UI to port 9080, change this value to: "9080:80".
      4. Save and close the file. When AXON Predict is running, the UI will be available at localhost:9080.

Starting AXON Predict

  1. Open a terminal window at the compose directory location.
  2. Log in to the Greenwave Systems Docker Registry with the account (username and password) you created when registering with the Edge Developer Portal:
    1. Type (or copy/paste) the following command:
      docker login

      (From a Microsoft Windows machine, you may have to use the following command instead:
      winpty docker login

    2. When prompted, type your username and press Enter.
    3. Then, type your password and press Enter. See the message, Login Succeeded. (Note: If you can’t remember your account credentials or are having problems logging in, register again to create a different account with new credentials.)

    Now run the Docker Compose commands to start Predict Hub (cloud) and then Predict Atom, in this order.

  3. Run the following:
    docker-compose -f cloud-compose.yml -p predict up -d --remove-orphans

    Compose starts the Predict Hub. You can open another command prompt and view the container log.

    Now you start the edge application, Predict Atom. Compose will also start the CSV-Adapter tool.

  4. Run the following:
    docker-compose -f atom-compose.yml -p predict-edge up -d --remove-orphans

    Compose starts Predict Atom and then CSV-Adapter. Again, you can open other command prompts to view the logs for Atom and CSV-Adapter.

    AXON Predict is now running on your machine. You can browse your localhost to see the AXON Predict UI with data provided by the real-time data simulator. The Quick Starts will guide you through a real-world scenario that uses this data for analytics, visualizations, and more.

Quick Starts

Ready to get started? The Quick Starts introduce you to the visualization tools and teach you how to create your own analytics so you can get valuable insight into your devices and processes in no time.