Enabling AI features

You enable AI features by activating AI configurations in the AI configuration application.

1. Enable AI insights for asset condition and alerts.
   Get AI-generated insights into assets and alerts that analyze asset health, maintenance history, and failure patterns to help you make faster, more informed decisions about asset maintenance and risk mitigation. Alert insights automatically provide context and recommendations when alerts are triggered, which can reduce response time and improve troubleshooting efficiency.
   AI insights for assets is available in 9.2. For more information, see Enabling AI insights for assets and alerts.
   1. In the AI configuration application, click Add configuration.
   2. Specify a name and, optionally, a description.
   3. In the Template field, select the insightsgenerator template.
   4. Select the latest template version.
   5. Click Create and then open the configuration.
   6. Click Actions > Activate.
   7. Wait a few minutes and then click Actions > Create model.
   8. Verify that the AI features are ready to use.
      • Asset insights are located on assets dashboards. From the Assets & locations dashboard, select an asset that has at least 5 work orders, locate the Insights card, and click Generate insights.
      • Alert insights are located in alert-based tickets. From the Tickets application, open a recent alert-based ticket that's associated with an asset. The Alert insights card is located on the Summary tab.
2. Enable recommended problem codes in Work orders.
   Speed up work order creation and improve data consistency by getting AI-recommended problem codes based on work order descriptions and other related data. Reduce manual data entry errors and help technicians quickly categorize work for better reporting and analysis.
   For more information, see Preparing your data for problem code recommendations and Enabling recommended problem codes in Work orders.
   1. Choose at least 20 work orders for the training data set. By default, the work orders must be EM or CM type work orders. 10 of those work orders must contain one problem code. The other 10 work orders must contain another problem code. Choose as unique, diverse, and accurate work orders as possible. In Maximo Manage, open Work orders, open a work order and then in the Actions menu, click Add to AI training model. A tag appears in the work order to indicate that the work order is added to the filter. Ensure that you save the work order. You can retrain the model later on new data.
   2. In the AI configuration application, select the WOPROBLEMCODE configuration.
   3. Click Actions > Check data requirement.
      When you check the data requirements, the training data is reviewed to determine whether the data contains enough problem codes and work orders. If the data check fails, you must add or improve the quality of data in your training filter. For more information, see Preparing your data for problem code recommendations.
   4. Click Actions > Activate.
   5. Click Actions > Train model and monitor the model status.
   6. In Work orders, find a CM or EM type work order that is in the WAPPR status, was reported less than 30 days ago, and is not in the training filter. Open the work order and in the Actions menu, click Edit. Verify that problem code recommendations are available for the Problem code field.
3. Enable locating of similar work orders.
   Use AI to find work orders with similar issues and solutions. Reduce troubleshooting time and helping your team learn from past experiences and resolve problems faster.
   For more information, see Enabling locating of similar work orders and Finding other types of similar data.
   1. In the AI configuration application, select the WOSIMILARITY AI configuration.
   2. Click Actions > Activate.
   3. Click Actions > Create model.
   4. Verify that the AI feature is ready to use.
      In Work orders, find a PM type or unidentified type of work order that is not cancelled or closed. Open the work order and in the Actions menu, click View similar work orders.
4. Enable AI recommendations in Reliability Strategies.
   Accelerate reliability strategy development with AI-generated recommendations for failure modes, effects, and mitigation actions. Help reliability engineers build comprehensive FMEA strategies faster and with greater confidence.
   For more information, see Enabling AI recommendations in Reliability Strategies.
   1. In the AI configuration application, open the AI configuration for the fmea template.
   2. Click Actions > Activate.
   3. Click Actions > Create model.
   4. Verify that the AI feature is ready to use.
      In Reliability Strategies, select a custom strategy and open the edit view or click Create strategy. AI recommendations are available for the following content:

      • On the Overview tab, in the Boundary field
      • On the Failure modes tab, from the hierarchy view
      • If you are using 9.2, on the Actions tab, for generating action descriptions
5. Enable an agentic AI assistant.
   Empower your users with an agentic and conversational AI assistant that can retrieve data, search documentation, generate insights, and complete tasks on their behalf using natural language. The assistant reduces training time, improves productivity, and makes Maximo more accessible to users of all skill levels.
   The agentic AI assistant is available in 9.2. The agentic Maximo Assistant requires more storage than the non-agentic assistant that was available in Maximo Application Suite 9.1. For more information, see Enabling the agentic Maximo Assistant.
   For more information about enabling the AI assistant in 9.1, see Enabling the assistant to retrieve data.
   1. From the navigation menu, click Administration > Configurations.
   2. Open the AI Service configuration, select the Deploy AI agent services checkbox, and save the configuration.
      Wait until the required deployments are running and ready.
   3. Set up the data retrieval tool and the docsearch tool. The data retrieval tool enables the assistant to find and retrieve data for you. The docsearch tool enables you to search for information about Maximo in the official IBM Documentation. If it is activated, the assistant will also use the insightsgenerator AI configuration as a tool.
      1. Open the ASSISTANT AI configuration.
      2. On the Searchable data tab, determine which object structures the assistant must access. You can ask questions about these objects after the assistant is enabled, and you can also add or remove objects from the list later.
      3. Click Actions > Activate.
      4. Click Actions > Create model.
      5. Open the DOCSEARCH AI configuration.
      6. Click Actions > Activate.
   4. Wait at least 10 minutes for the MCP server to refresh with the newly activated tools.
      The MCP server refreshes available tools every 10 minutes. If you need the tools to be available immediately, you can force a refresh by deleting the MCP server pod in Red Hat® OpenShift® web console. In your Maximo Manage project, locate the instance_id-mcp-string pod and delete it. When the pod restarts, the tools are immediately available.
   5. Return to the AI configurations table and on the Maximo Assistant tile, click Configure.
   6. Enable the Agentic assistant toggle.
   7. In the tools table, verify that the nl2oslc and docsearch templates are listed.
      Maximo Assistant answers prompts based on any available tools or, at its discretion, its general knowledge. Maximo Assistant can decline to answer a prompt if it determines the prompt does not relate to asset lifecycle management.
   8. Click Save.
   9. Verify that the assistant is ready to use.
      Open Maximo Assistant from the AI icon and test prompts that requires the assistant to retrieve data, search IBM Documentation, or get insights.

      1. Get all open work orders at site
      2. How do I create a work order in Maximo Manage 9.2?
      3. Generate insights for asset number