Georgia Pacific Transforms Manufacturing with C3 AI on AWS

GP is one of the world’s leading makers of paper, tissue, packaging, and building products, with over 100 manufacturing facilities across North America. Their paper mill manufacturing facilities are large-scale, multi-process operations producing thousands of tons of paper per day. Monitoring the status of just one machine can easily involve over 5,000 sensors that produce billions of data records daily.

With goals of achieving optimal manufacturing conditions, maximizing the life of their critical assets, including boilers, pumps, turbines, etc., in a sustainable and environmentally conscious manner, GP sought an AI solution that would support continuous operations monitoring to help prevent asset downtime. Having already migrated some internal systems to Amazon Web Services (AWS) to reduce data center costs and previously created a cloud-based analytics solution that leveraged multiple AWS services like Amazon Kinesis, Amazon Simple Storage Service S3, Amazon EMR, and Amazon SageMaker to optimize processes, GP scouted four AWS Partners with a solution that could help take analytics even further while integrating many disparate data sources at scale.

To accomplish its goals, GP partnered with C3 AI to embark on a comprehensive, multi-year partnership designed to expand the data integration and advanced AI capabilities of their Collaboration and Support Center (CSC) an internal team at GP established to improve operational reliability drastically. Initial monitoring implemented by the CSC was effective for simple assets with single tags, but it could not scale to accommodate complex assets that included multiple processes, components, and sensors. This resulted in the CSC being unable to support ML model pipelines that had multiple tags and additional data sources.

At the outset, GP and C3 AI established a Center of Excellence (CoE) with members from both companies to support identifying, prioritizing, developing, and deploying AI-based solutions. One of the first projects identified was the opportunity to use the C3 AI Platform and available applications for improved operations monitoring and reducing unplanned downtime for complex assets and other critical equipment.

By deploying within GP’s AWS environment, the CSC experiences all the benefits of AWS: security, scalability, cost-effectiveness, and cloud-native solutions. With a data center and network architecture built by AWS to meet the requirements of the most security-sensitive organizations, they can protect sensitive proprietary data. This process has laid the foundation to scale from 60 to more than 200 of these assets and introduce new asset classes into the C3 AI Reliability app.

The C3 AI Reliability application was deployed for a subset of critical assets across facilities, which required unifying sources of process data, vibration data, downtime indicator data, work orders, and asset hierarchies into the application. Rich and integrated application workflows help CSC engineers investigate alerts and collaborate with facilities more effectively to resolve risks. By taking a more holistic approach to monitoring complex assets—especially with large-scale data integration on AWS—the CSC has streamlined performance monitoring and significantly reduced the number of alerts generated.

Today, Georgia-Pacific has rolled out C3 AI Reliability to monitor over 200 large assets across facilities and plans to expand the application’s use.

With the combination of more data, contextualized sensor readings, and advanced ML techniques, GP has seen up to 5% improvement in overall equipment effectiveness (OEE) from their C3 AI Reliability deployment. With the continued initial success, GP plans to expand the scope of C3 AI Reliability to eight additional asset classes in the next year.