Reducing Mean Time to Repair (MTTR) with Amazon Q Business

Unplanned downtime can cost consumer-goods manufacturers millions each year. Production-line disruptions often create ripple effects throughout supply chains, reducing inventory and negatively affecting the customer experience. While advancements in monitoring and predictive maintenance solutions have helped manufacturers reduce downtime, they still need to act quickly when an unplanned disruption occurs. Manufacturers also need solutions to optimize planned maintenance events. In this blog post, we’ll analyze a typical technician’s workflow, identifying opportunities to use Amazon Q Business in each phase to reduce Mean Time to Repair (MTTR).

Amazon Q Business can streamline tasks and accelerate problem solving in resolution workflows. This allows the technician to operate more efficiently. They’ll spend less time seeking the correct information, issuing commands to obtain data or manage equipment, writing documentation, and following up. Improving the accuracy of the information available to technicians should also improve metrics on interruptions.

Challenges with traditional approaches to handling disruptions

Traditional methods of handling production-line disruptions are often slow and inefficient. Technicians must navigate through multiple phases—preparation, diagnostic, execution, and completion to identify and resolve issues. Throughout this process, technicians often lack real-time access to the critical information they need. To find answers and document work, they must manually search through extensive documentation, such as manuals, standard operating procedures (SOPs), and safety procedure guidelines.

The typical workflow of a shop-floor technician involves resolving a production-line disruption or completing a maintenance task. First, the technician goes through the preparation phase, reviewing daily work orders and assignments. Next, in the diagnostic phase, the technician needs to inspect the equipment, run an initial test, determine the required steps for resolution, and document the equipment’s current state. In the work-order execution phase, the technician needs to order parts, ensure compliance with all operating and safety procedures, and perform required repairs or maintenance. Finally, in the completion phase, the technician runs quality and safety tests, documenting findings throughout.

Figure 1: The typical workflow of a shop-floor technician when resolving a production-line disruption or completing a maintenance task.

Amazon Q Business transforms how technicians can respond to disruptions

Amazon Q Business is a fully managed, generative-AI powered assistant. Configure it to answer questions, provide summaries, generate content, and complete tasks. End users receive immediate, permissions-aware responses from enterprise data sources with citations such as internal documentation, operating procedures, and safety guidelines. Amazon Q Business allows technicians to access information quickly and streamline each phase of the maintenance workflow.

During each phase of the repair process for a shop-floor machine, technicians can use Amazon Q Business to accelerate repairs, as follows:

• Preparation phase—Receive summarized access to work orders and relevant maintenance history.
• Diagnostic phase—Quickly pull documentation and SOPs for the work order. Analyze operational metrics using Amazon Q Business with a connector to Amazon QuickSight, a cloud-powered, serverless business intelligence service that allows users to build interactive dashboards, perform ad-hoc analysis, and generate insights. Run real-time diagnostics on the equipment using custom plugin functions from Amazon Q Business. Efficiently identify the proper corrective actions by comparing historical and real-time maintenance data and documentation.
• Execution phase—Order parts or initiate a software-based reset procedure.
• Completion phase—Run additional tests and use telemetry and metrics to generate case-closure reports. Schedule a follow-up visit.

During all phases, Amazon Q Business operates securely and respects access control rules by integrating with corporate identity providers. This ensures that only authorized technicians have access to the information they need to perform the appropriate actions.

The time required in each phase is highly dependent on the specific problem. However, the time spent interacting with multiple systems and gathering information for the repair can be significantly reduced by using integrations provided by Amazon Q Business. For example, McKinsey found that with generative AI, maintenance labor costs have dropped by a third. Additionally, technicians now have 40 percent more capacity, since they rarely need to answer simple questions from operators or help resolve routine breakdowns.

Solution overview

The following solution architecture (Figure 2) is a guideline for using Amazon Q Business to improve MTTR. The diagram explains how Amazon Q Business uses its connectors and plugins to integrate with other services. Amazon Q Business integrates seamlessly with existing systems like Amazon S3, which can be used as a repository for documentation and SOPs, with Jira for ticket management. It can integrate with AWS Lambda for custom document enrichment, extracting metadata from the images and text, processing the metadata in Amazon Bedrock, and using Amazon QuickSight for metrics.

With Amazon Q Business custom plugins, you can integrate with virtually any third-party application, like APIs and a Manufacturing Execution System (MES). Secure access controls and guardrails ensure that only authorized users have access to content or can request Amazon Q Business to perform an action. Additionally, Amazon Q Apps allow the technician to further improve this solution by creating custom apps that streamline their tasks, helping to improve prompts and reduce the details needed.

Figure 2: This architecture diagram explains how Amazon Q Business uses its connectors and plugins to integrate with other services.

This solution is especially relevant to Consumer Packaged Goods (CPG) customers because they use highly automated supply chain operations, but it can also scale across industries including telecommunications, energy, retail, QSR, and medical equipment. We recommend starting with a pilot implementation to demonstrate ROI through reduced downtime costs and improved operational efficiency. To get started, please review Getting started with Amazon Q Business.

Solution in action: streamlining factory operations with Amazon Q Business

This solution demonstrates how Amazon Q Business transforms daily operations for factory technicians through a unified interface called MaintainIQ. Using their existing identity provider, the technician can access a powerful platform that consolidates multiple systems and workflows into a single conversational interface.

Figure 3: Amazon Q Business web interface for MaintainIQ; AI assistant for shop floor created for this demo.

In our demonstration, we followed a typical workflow where a factory worker needed to troubleshoot a fan issue (Fan T200) on a test device. Using the Amazon Q Business Jira integration, the worker could instantly view their open tasks.

Figure 4: Amazon Q Business web interface showing the list of open Jira issues assigned to the technician.

When needing to access SOPs, instead of searching through hundreds of pages of documentation, they can simply ask Amazon Q Business how to handle the fan replacement, receiving immediate, relevant information from the knowledge base. Notice that SOPs and other documentation are referenced after each recombination, with a card indicating the reference number.

Figure 5: Amazon Q Business web interface showing SOPs related to the solution of the issue assigned to the technician.

Through the Amazon QuickSight plugin, workers can analyze real-time fan speed data without leaving the interface. As we can see in the following image, the speed is dropping and as per SOP guidance, the technician should update it to 1500 RPM in MES.

Figure 6: Amazon Q Business web interface showing integration with Amazon QuickSight for near real-time equipment monitoring.

With the Amazon Q Business custom plugin created for MES, the technician can control factory equipment directly through the Amazon Q Business interface. When the fan required replacement, the Jira plugin facilitated immediate creation of a purchase order ticket—all within the same conversational interface.

Figure 7: Amazon Q Business web interface showing a custom plugin for integration with MES and near real-time operations on equipment.

Once Amazon Q Business completes the request to update, the updated MES response is “FAN T200.”

Figure 8: Amazon Q Business web interface showing a custom plugin for integration with MES and near real-time equipment telemetry.

Then, according to the SOP from Amazon Q, the technician requests to query Amazon QuickSight for fan speed again. This shows that the fan speed of T200 continues to drop after a few minutes. Hence, according to the SOP, the technician should replace the fan. Using the Jira plugin, the technician requests Amazon Q to create an issue in Jira requesting a purchase order to be created. The technician can then review the information entered in the issue and submit.

Figure 9: Amazon Q Business web interface showing a Jira plugin to create an issue for a part replacement and purchase order.

This implementation showcases how Amazon Q Business eliminates the need to switch between multiple systems, significantly improving efficiency and reducing the complexity of daily factory operations. Combining documentation access, data analysis, equipment control, and workflow management into one interface allows customers to deliver a streamlined and accessible working environment for factory personnel.

Conclusion

Amazon Q Business has the potential to revolutionize how manufacturers handle production-line disruptions. Using the power of generative AI, manufacturers can reduce MTTR, improve service quality, increase operational efficiency, and deliver exceptional value to their teams and customers. In this post, we identified key opportunities to create immediate value for CPG customers across each phase of shop-floor machine repair. Additionally, Amazon Q Business can analyze operational metrics using Amazon QuickSight connector. It can run real-time diagnostics on equipment, using custom plugin functions, and compare historical and real-time data with maintenance history within context, providing documentation to efficiently identify the proper corrective actions.