Från variation till standard: En studie om standardisering under produktionsupptrappning

Sammanfattning

This study investigates how variation manifests in manual processes during production rampup, and how standardized work instructions should be designed to address it. The study was conducted as a case study at a Swedish manufacturing company in transition from prototype to series production. Data was collected over 72 working days through a combination of quantitative production yield data and qualitative methods including semi-structured interviews with 14 operators and Gemba Walk observations. The analytical framework applied is DMAIC, supported by Pareto analysis and Ishikawa diagrams. The study identifies three forms of variation: hidden variation, where defects are registered downstream from their origin; operator-dependent variation, correlated with shift changes rather than technical process parameters; and behavioral variation driven by information gaps that generate material waste. A central finding is a measurement paradox in which conventional quality metrics such as Pareto analysis can be systematically misleading during ramp-up. A station may display near-perfect yield figures while simultaneously being the primary source of defects in the production flow. The study concludes that variation during ramp-up is primarily an information problem rather than a competence problem. In response, a standardized work instruction template was developed with four empirically motivated additions beyond existing literature: flow context, material rules, visual approval criteria, and a validation field. Standardization is argued to be not only a means of reducing variation, but a prerequisite for locating it.