Get Started#

  • Time to Complete: 30 minutes

  • Programming Language: Python 3

Prerequisites#

Configure Docker#

To configure Docker:

  1. Run Docker as Non-Root: Follow the steps in Manage Docker as a non-root user.

  2. Configure Proxy (if required):

    • Set up proxy settings for Docker client and containers as described in Docker Proxy Configuration.

    • Example ~/.docker/config.json:

      {
        "proxies": {
          "default": {
            "httpProxy": "http://<proxy_server>:<proxy_port>",
            "httpsProxy": "http://<proxy_server>:<proxy_port>",
            "noProxy": "127.0.0.1,localhost"
          }
        }
      }
      
    • Configure the Docker daemon proxy as per Systemd Unit File.

  3. Enable Log Rotation:

    • Add the following configuration to /etc/docker/daemon.json:

      {
        "log-driver": "json-file",
        "log-opts": {
          "max-size": "10m",
          "max-file": "5"
        }
      }
      
    • Reload and restart Docker:

      sudo systemctl daemon-reload
      sudo systemctl restart docker
      

Clone source code#

Go to the target directory of your choice and clone the suite. If you want to clone a specific release branch, replace main with the desired tag. To learn more on partial cloning, check the Repository Cloning guide.

git clone --filter=blob:none --sparse --branch release-2026.3.0 https://github.com/open-edge-platform/manufacturing-ai-suite.git
cd manufacturing-ai-suite
git sparse-checkout set industrial-edge-insights-time-series
cd industrial-edge-insights-time-series

Deploy with Docker Compose#

  1. Update the following fields in .env:

    • INFLUXDB_USERNAME

    • INFLUXDB_PASSWORD

    • VISUALIZER_GRAFANA_USER

    • VISUALIZER_GRAFANA_PASSWORD

  2. Deploy the sample app, use only one of the following options:

Note

  • The below make up_opcua_ingestion or make up_mqtt_ingestion fails if the above required fields are not populated as per the rules called out in .env file.

  • The sample app is deployed by pulling the pre-built container images of the sample app from the docker hub OR from the internal container registry (login to the docker registry from cli and configure DOCKER_REGISTRY env variable in .env file at manufacturing-ai-suite/industrial-edge-insights-time-series)

  • The CONTINUOUS_SIMULATOR_INGESTION variable in the .env file is set to true by default, enabling continuous looping of simulator data. To ingest the simulator data only once (without looping), set this variable to false.

  • If CONTINUOUS_SIMULATOR_INGESTION is set to false, you may see the [inputs.opcua] status not OK for node message in the telegraf logs for OPC-UA ingestion after a single data ingestion loop. This message can be ignored.

  • make up_opcua_ingestion is supported only for Wind Turbine Anomaly Detection sample app

  • Using OPC-UA ingestion:

    make up_opcua_ingestion
    
  • Using MQTT ingestion:

    make up_mqtt_ingestion
    

Running User Defined Function(UDF) inference on GPU#

By default, UDF for Wind Turbine Anomaly Detection sample app is configured to run on CPU.

To trigger the UDF inference on GPU in Time Series Analytics Microservice, run the following command:

cd industrial-edge-insights-time-series/apps/wind-turbine-anomaly-detection/time-series-analytics-config
curl -k -X 'POST' \
 'https://127.0.0.1:3000/ts-api/config' \
 -H 'accept: application/json' \
 -H 'Content-Type: application/json' \
 -d "$(sed 's/"device": "CPU"/"device": "GPU"/' config.json)"

Verify the Output Results#

  1. Get into the InfluxDB* container:

     docker exec -it ia-influxdb bash
    
  2. Run the following commands to see the data in InfluxDB*:

    Note

    Please ignore the error message There was an error writing history file: open /.influx_history: read-only file system happening in the InfluxDB shell. This does not affect any functionality while working with the InfluxDB commands

    # For below command, the INFLUXDB_USERNAME and INFLUXDB_PASSWORD needs to be fetched from `.env` file
    influx -username <username> -password <passwd>
    use datain # database access
    show measurements
    # Run below query to check and output measurement processed
    # by Time Series Analytics microservice
    select * from "wind-turbine-anomaly-data"
    
  3. To check the output in Grafana:

    • Use link https://localhost:3000/ to launch Grafana from browser (preferably Chrome browser)

      Note

      • If you are accessing the UI remotely, replace localhost with the host system IP address.

    • Login to the Grafana with values set for VISUALIZER_GRAFANA_USER and VISUALIZER_GRAFANA_PASSWORD in .env file.

      Grafana login

    • After login, click on Dashboard Menu view

    • Select the Wind Turbine Dashboard. Windturbine dashboard

    • You will see the below output.

      Anomaly prediction in grid active power

Bring down the sample app#

make down

Check logs - troubleshooting#

Check container logs to catch any failures:

docker ps
docker logs -f <container_name>
docker logs -f <container_name> | grep -i error

Advanced setup#