Self-resilience production systems: Six sigma fault diagnostics and process adjustment in closed loop lifecycle

Manufacturing firms are perpetually aiming to reduce their internal and external lead time to launch new products as a way to respond to ever-changing customer needs. In order to stay competitive and get the highest possible return from capital equipment, industries are using modular design of products as well as flexible, reconfigurable, and reusable processes to provide necessary variety in their product line for the end-customer. These strategies put tremendous pressure on dimensional quality due to reduced lot sizes (production batches) and the increasing varieties of products produced on the same line.This study aims to develop a self-resilience framework for production system to systematically monitor, diagnose and adjust process to eliminate six-sigma faults during the manufacturing process. The study addresses two specific issues: first, the diagnostics of six-sigma faults related to dimensional quality and next, process adjustment to correct the root cause(s) of the faults or analysis of process capability to recover from the fault.In this study, a diagnostics methodology for six-sigma faults for compliant parts assembly is developed. The increasing complexity of modern products coupled with part compliance/flexibility frequently results in ill-conditioned assembly systems further adding to the challenges of process control and fault failure diagnosis. The study presents a model-based approach using statistical modal analysis (SMA) and data-based approach using Enhanced Piecewise Least Square (EPLS) approach for isolating dimensional six-sigma fault failure in ill-conditioned single-station and multi-station assembly systems.Current advances in reconfigurability and reusability based on modular design have enabled the assembly of product family on a single assembly line. However, these strategies make adjustments and calibration of assembly fixtures much more complex due to the sharing of tooling between products. A model-based process adjustment approach is proposed to eliminate assembly fixture-induced product failure in multi-station assembly system that produces product family on a parallel or serial assembly line configuration.With respect to all the abovementioned research topics, corresponding case studies from automotive assembly processes are provided to demonstrate the presented methodologies.