Q3 Technologies developed a Predictive Maintenance Solution for one of the world’s largest telecom equipment manufacturer, to predict the failure of machines 24 hours in advance. The client wanted to reduce defects caused in production line from 8-9 defects per million to ~3.4 defects/million.
Q3 Tech Team developed the solution where the client can see the real-time dashboards having details about all the machines in the production line and about the defects on each of the machines. This helped the client to reduce defects significantly by taking corrective actions on defective machines.
Technology Used – LSTM (Long Short Term Memory), RNN (Recurrent Neural
Business Benefits: Anticipated & Achieved
- Reduction In no. of defects from 8-9 defects/Mn to ~3.4 defects/Mn products
- Improved Maintenance Schedule By predicting failures in machines 24 hrs. in advance
- Improved Production with Lower rate of Machine Failures
- Costs Saving on Maintenance Schedule & Defects