AkzoNobel is a Dutch multinational company that produces performance coatings for industries and consumers worldwide. The modern-day AkzoNobel came about in 1994 through the merger of the Dutch Akzo and the Swedish Nobel Industries, creating one of the leading coatings companies in the world.

AkzoNobel’s North American production sites focus on manufacturing plastic powder coatings that can be sprayed directly onto industrial surfaces for a resilient finish. To make these coatings, operators manually measure out raw materials according to each coating’s strict formula. This manual measurement is a source of human error, and an incorrect dosage may lead to significant production losses.

Our AI-Enabled Raw Material Dosing Accuracy Monitoring system assists AkzoNobel operators in discovering dosing errors early. Our software monitors dosing errors through cameras and provides real-time alerts when the wrong material or quantity is added. Additionally, these alerts are logged to identify key points in the manufacturing process that are most susceptible to issues.

The operator dashboard focuses on simplicity of use, with a provision to scan material lists. An alert display shows deviations and previous failures for analysis. Additionally, a tablet interface is available at every dosing station to detect dosing errors in real time, with review systems to mitigate the errors.

All actions are recorded for future review for quality control concerns, traceability and efficient resolution of pipeline issues. This enables efficient tracking of the manufacturing process without hindrance to the operators performing the dosage.

Our back end is built in Python using PaddleOCR and OpenAI’s GPT-6 Luna for computer vision, TypeSafe AI for accuracy analysis and MySQL as our database. The front end is implemented in React and React Native written in TypeScript and CSS.