Failure Diagnosis Using Decision Tree
White Papers
Failures in manufacturing process occurs when function cannot be performed or has performances that deviate from target performance. Failures can cause damage and affects operations, productions, performance, and further impact company brand and reputation. There are few categories for failures:
• Aging faults
• Installation faults
• Software faults
• Design faults
• Operational faults
• Excessive stress faults
• Other reasons
Typical failure diagnosis is to conduct root causes analysis like Fishbone Diagram, 5 Whys, 8 principles, etc. However, these analyses are laborious and time-consuming, it involves SME (subject matter experts), interviews, process improvement teams, lean six sigma experts, etc. A data driven decision by a big data analytic platform has advantages in time-to-inference and continuous improvement compared to traditional failure diagnosis approaches.