REDUCING LABOR CONTENT AS A STRATEGY TO IMPROVE COMPETITIVENESS - AN ANALYSIS THAT ADDRESSES THE VALUE OF DESIGNING FOR AUTOMATION AND AN EMPIRICAL ANALYSIS THAT EXPLOITS THE AUTOMATION USING META PROCESS CONTROL

摘要

Reducing the labor CONTENT of an electronic product assembly has been shown to be an effective strategy to reduce labor cost. In many cases it permits production operations in high labor rate markets to successfully compete with operations that have available sources of local low cost labor. There have been a number of papers written to support and justify this strategy. Automating assembly processes that are traditionally manual is the basis of a tactical plan that potentially allows the strategic objective of reducing labor content to be met. The word potentially is used with purpose. Investing in hundreds of thousands of dollars of leading-edge automation equipment does not an automated factory necessarily make. If the automated processes require constant human monitoring and intervention, or electrical test yields are low and rework labor high, the objective of the automation (and the capital investment) is defeated - we say instead of automating, we faux automate. The objectives of this paper are to: 1. Provide examples of real world automation opportunities for future conversion from high to low labor content and, 2. Address the key elements that are needed to exploit this automation - not simply faux automate. Specifically: a. Designing a product for automated assembly. b. A factory infrastructure that has proactive process control built in and, ideally, automated proactive process control that self-corrects process issues, i.e., Meta Process Control. c. A skilled workforce that can develop and maintain the automated processes organized in a way to minimize cost. To achieve these objectives the work that this paper documents was divided into two distinct parts: 1. A case study undertaken to identify and analyze two opportunities that can use automation to reduce the labor cost of a product, i.e., reduce the labor content by converting to or initially designing for assembly automation. 2. An initial experiment that establishes the process window associated with printing solder paste for extremely small components (01005 English / 0402 Metric), i.e., establish the need to proactively monitor process parameters and ideally to auto-correct, i.e., Meta Process Control (MPC), to fully exploit the value of automating for a process with an extremely small process window. The results verify that: 1. Designing for automation can be an effective tool in eliminating the competitive advantage that low labor rate operations provide. In the opportunities identified, the total labor cost of a product assembled in a high labor rate region can be dramatically reduced by automating labor intensive hand soldering operations. Further, if the product is initially designed by combining the added automation with existing process assembly automation equipment, payback on investment will be accelerated. The paper predicts that based on a real world production model, automating two labor intensive assembly processes would result in a reduced labor cost of about 45% with a return on capital investment of about 14 months for the one process (11 months if a product redesign is not required). The coax soldering ROI is TBD. The results of actually automating these opportunities will be documented in a future paper. 2. Achieving low yield loss from any automated process involved in the assembly of the product (including electrical test false positives) is essential. The experiment that was designed used a component with a narrow process window to demonstrate the futility and increased cost in automating without an accompanying capable and controllable process. Further, simply identifying a process that has violated its process window is not enough. This has traditionally been the case with failures found at In-circuit test (ICT). By this time, there are typically many circuit boards that require expensive troubleshooting, rework and retest in a high labor rate environment. The experiment demonstrated that true test yields above 95% are required to take