In modern manufacturing, quick access to critical information and access to the expertise of experienced operators are essential to maintaining high productivity and quality and minimizing costly machine downtime. However, as skilled workers retire or transfer and the workforce landscape changes, manufacturers often face challenges in preserving knowledge within their organization and smoothly transferring it to new and junior employees. This is important to address because according to Deloitte research, 2.69 million manufacturing jobs are expected to become vacant due to retirements, and an additional 1.96 million new jobs will be created through natural growth. More worryingly, it states that 53% of open positions may go unfilled due to skills shortages in manufacturing, highlighting the urgent need for effective knowledge transfer and preservation strategies.
Executives at Georgia-Pacific, a leading manufacturer of tissue, pulp, paper, packaging, and construction products, recognized that they were facing the same challenges that other manufacturers were struggling with. That means employees spend hours searching for troubleshooting information when a problem occurs. We needed to transform our approach to knowledge management so that critical information could be accessed within seconds instead of hours. Using Amazon Web Services’ generative AI capabilities, we built a centralized knowledge hub that gives operators instant access to the guidance they need. The solution helped the team quickly resolve production issues, reduce material defects, and minimize waste.
This blog describes how to implement generative AI solutions using Amazon Bedrock, a comprehensive, secure, and flexible platform for building generative AI applications and agents. This solution helps manufacturing companies accelerate operator migration and onboarding, gain the expertise of skilled employees, and minimize machine downtime across vast manufacturing operations.
the gap we bridge
Many manufacturers struggle with time-consuming research and discovery across disparate knowledge systems when searching for information. This is especially difficult in industries with complex systems and maintenance requirements, where the departure of one experienced technician can immediately cause operational disruption.
As technology advances accelerate and workforce demographics change, organizations must prioritize systematic knowledge transfer programs that bridge the gap to maximize operational impact and maintain business productivity.
Some of the key challenges facing manufacturing companies include:
- Lack of centralized and accessible knowledge: The lack of a readily available, centralized knowledge base makes it difficult for employees to quickly find the right information when a problem arises. Employees must search multiple systems, contact experts, and review physical documents, all of which can delay problem resolution. This leads to reduced machine productivity, extended downtime, and increased troubleshooting and repair costs throughout your manufacturing facility.
- Risk of losing institutional knowledge: Manufacturers face the risk of losing valuable organizational knowledge as experienced operators who have been running production lines for decades retire or move on.
- Inadequacies of traditional knowledge management approaches: Relying on calls to experts or searching for physical documents is insufficient in a fast-paced, data-driven manufacturing environment and often increases issue resolution times and increases the risk of downtime.
Manufacturers need scalable and effective solutions that integrate disparate information, make information accessible to machine operators, preserve the expertise of the most knowledgeable employees, protect organizational knowledge, and improve operational efficiency.
Solution overview
Georgia-Pacific tackled these manufacturing challenges by creating an AI-powered assistant called ChatGP on Amazon Bedrock. Together with AWS, they built a system that combines a chatbot built using a cloud-based model via Amazon Bedrock with real-time machine data to help operators solve problems and improve production.
Georgia-Pacific’s team recorded conversations with experienced workers and subject matter experts about older equipment that lacked proper documentation. The team used Amazon Bedrock to transform these discussions into structured technical documentation. This allowed tribal knowledge that previously existed only in the heads of veteran employees to be preserved for decades.
The system assists operators in real-time by providing step-by-step guidance on machine adjustments and troubleshooting. Because the system is connected to the machine’s sensors, it combines past knowledge with current operating conditions to provide more accurate advice. Operators access it through a simple web interface from any computer or tablet on the network.
Georgia-Pacific’s success with ChatGP shows the potential of AI-powered knowledge management in manufacturing environments. The solutions outlined below can help manufacturers build similar knowledge management systems that use chatbots to gather organizational knowledge from experienced operators, provide instant access to troubleshooting guidance, and accelerate new employee onboarding. This approach enables organizations to preserve decades of tribal knowledge while providing employees with immediate, contextual answers to complex operational challenges, transforming knowledge transfer from time-consuming manual processes to efficient, scalable digital solutions.
solution architecture
Figure 1: High-level solution architecture for knowledge management chat bot assistant
This solution provides an overview of how to implement a knowledge management chatbot assistant using Amazon Bedrock. The solution is divided into four main steps.
- acquisition of knowledge: A senior machinist records machine-related information through a web interface and then uploads it to Amazon Simple Storage Service (Amazon S3), an object storage service that serves as the primary repository for all information assets.
- automatic transcription: Amazon Transcribe converts uploaded recordings from audio to text, making the expertise of advanced mechanics searchable and accessible in written form.
- Creating a knowledge base: The transcribed text serves as the data source for Amazon Bedrock’s knowledge base and forms the basis of a generative, AI-powered knowledge hub.
- Intelligent response generation: Junior mechanics can access a Q/A chatbot interface to ask questions about machine operation, troubleshooting, and maintenance in simple, natural language. When a junior machinist submits a query, it leverages Amazon Bedrock text generation language models in combination with a knowledge base to power search augmented generation (RAG) to provide relevant and contextual responses.
In Figures 2 and 3, we share videos that demonstrate the two main workflows of this solution in action. The first workflow, shown in Figure 2, shows how a mechanic can create and upload a record. Once uploaded, the system automatically converts the recording to text and stores it as a knowledge base within Amazon Bedrock, as shown in Figure 3. The second workflow shows how mechanics can easily troubleshoot problems by asking questions using simple natural language prompts.
Figure 2: Machine Operator Assistant Workflow 1
Figure 3: Machine Operator Assistant Workflow 2
conclusion
By implementing an AI-powered knowledge hub using Amazon Bedrock, manufacturers reduce troubleshooting time from hours to seconds, retain organizational knowledge from departing workers, and accelerate operator onboarding.
Georgia-Pacific’s success with generative AI on AWS shows how manufacturers can address long-standing operational challenges. By consolidating disparate information into a centralized knowledge base, companies can reduce machine downtime, reduce repair costs, preserve critical expertise for future generations, and serve as a blueprint for operational excellence.
Discover our comprehensive manufacturing and industrial solutions in the AWS Solutions Library and work with your AWS account team to explore how AWS is driving manufacturing innovation. For more insight into the transformative impact of generative AI across product engineering, production optimization, and supply chains in manufacturing, read our blog How Generative AI will transform manufacturing.
