IoT – Predictive Modelling for the Paper Industry

Challenge

This global producer of process automation equipment supplies dryers to paper mills. Threshold breaches in vibration and moisture were leading to sheet breaks, so the company required a highly accurate predictive model.

Solution

TeraCode analyzed and modeled the data variables independently and interdependently, predicted the effects on the parameters (where possible), and extrapolated other parameters that could be impactful.

Technology

TeraCode’s data scientists built models testing several techniques: inherent pattern recognition, long and short term memory, and gradient boosting regression.

Outcome

The models achieved 99% accuracy up to an hour before a threshold breach occurred. The company connects models in real-time and displays results in a dashboard that prescribes preventative actions.

We are eager to help you on your digital transformation journey

Get in touch

Close menu