Forecasting quality in robotic controller supply chain

Abstract

Between 2017 and 2019, industrial robot installations are estimated to increase by 13% on average per year. As the industrial robot market has grown, so too have customer demands. Many industrial robot manufacturers are in a position to capture this growth by improving in on-time delivery, quality performance, and product offerings. This master's thesis is devoted to providing manufacturers methods for increasing quality performance for robotic controllers. To improve quality performance, we focus on finding the connection between controller quality performance at suppliers, manufacturing, and customer sites. We consider this valuable in the context of a manufacturer's R&D to set operational quality targets and predict the cascade effect in the supply chain. This study includes a deep dive in quality performance and methods to predict future performance. The analysis includes a look at quality metrics in the robot industry. We forecast future performance against these metrics using an available dataset and regression modeling. Because we do not discover strong regression models, we propose dataset statistics to forecast future quality performance. We have 2 recommendations based on our research. Moving forward, we recommend increasing transparency for quality data collection to create a more robust model with stronger prediction capabilities. We also recommend a total cost of quality approach in determining ideal quality metrics.