A leading manufacturer of flooring approached System One to develop an application and process that uses Artificial Intelligence (AI) to reduce or eliminate failures in their production process. System One built an AI model to perform Machine Learning. We created our largest model to date, with 7.5 billion data points used to create predictions in the manufacturing process of sheet flooring.
The client was having repeated problems in their manufacturing process. When a flaw was produced, it was repeated on an entire roll of vinyl flooring. This resulted in rolls of flooring being either scrapped or sold as a second in an attempt to recover costs.
System One Solution
System One undertook the project to build an overarching solution. There were three components to the solution:
- Artificial Intelligence in manufacturing, including hardware customization and deployment;
- Advanced Analytics Data Science and visualization;
- Industry 4.0 Smart Manufacturing.
Different types of sensors were used, including sensors capable of handling predictive maintenance, and others that could measure VOCs, CO2, ozone, temperature, humidity, and vibration.
System One developed an application, the Artificial Intelligence Hub, that uses AI to simulate a digital twin of a plant. With this application, the client can predict – with 87% accuracy – an impending scrapping event.
With these technological improvements, the client experienced invaluable results. Scrapping events were drastically reduced.