 |
The need:
complex manufacturing process spread across the world.
Production process malfunction may not be detected for
a couple of days resulting in large losses. On one such
occasion 46% of heads were found to be faulty.
The Project Objective: Identifying faults prior to testing,
saving the company millions. Efficacy of data mining in
predicting faults in the manufacturing process based on
data collected from the various stages of manufacturing.
The Technology: Decision trees and Neural networks,
Result: Successful detection of change in the production
environment, which had led to the increased error rate
in the production. |
