Challenge
Operators at a large manufacturing company are responsible for the performance and availability of large, complex machinery across manufacturing plants. Over the past few years, the company experienced increasing employee turnover driven largely by an aging workforce and pandemic-related disruptions. As a result, the company onboarded a wave of new operators who often struggled with troubleshooting the root cause identification of equipment issues.
With a distributed data footprint and relevant domain expertise concentrated with few experienced operators, the newly onboarded operators must go through a lengthy process to diagnose the root causes of failing equipment and determine corrective actions. When an operator observes an operational issue, he must either turn to experienced technicians or go through lengthy troubleshooting guides to determine potential root causes. The operator also reviews PI readings and compares historical values against desired values to identify out-of-bounds parameters. After the operator has identified potential causes, he typically goes back to the experienced technicians or consults troubleshooting guides and standard operating procedure documents to determine the appropriate action steps. The entire process is time-consuming and prone to human errors, hindering the company’s ability to rapidly react to issues and minimize unplanned downtime.
Approach
The manufacturing company decided to leverage a generative AI-powered knowledge assistant to empower the new operators and initiated a production deployment of C3 Generative AI across one of their major plants. The manufacturer chose to deploy C3 Generative AI with the solution’s enterprise access controls, deployment speed, and future scalability.
First, the manufacturing company and C3 AI team identified the relevant data types for improving operator productivity, such as standard operating procedures, troubleshooting guides, textbooks, job aids, and sensor data. The C3 AI team then unified data from 5 disparate sources including 3 document types and 2 process data types into an enterprise knowledge base. Using a proprietary large language model finetuned by C3 AI, the C3 AI team validated and tested the performance of the model against an identified library of relevant user queries.
With C3 Generative AI, the operator can utilize a simplified workflow to diagnose root causes and determine corrective actions. When operators observe a production issue, they can go directly to C3 Generative AI to search against troubleshooting guides and textbooks to identify potential causes. From the same simple interface, operators can query against sensor data to identify out-of-bound parameters. Finally, operators can search within operating procedures, troubleshooting guides, and job aids to identify corrective actions. The enhanced workflow helps operators improve productivity and their decision-making process.
Solution Architecture
Proven results in weeks, not years
Get insights into C3 AI’s capabilities, enterprise AI best practices, and highest-value use cases.
Gain insights into the C3 AI Platform's capabilities, its model-driven architecture, and test it against your company's sample data set.
Identify a high-impact business problem and collaborate with the C3 AI team to rapidly build an AI application that solves it.
Scale and deploy a tested C3 AI application into production. Incorporate user feedback and optimize algorithms to drive maximum economic value.