Crisplant defects quantification and reduction at an amazon.com distribution center

Abstract

Crisplant is a tilt-tray sortation system used in Reno (RNO 1) fulfillment center (FC) to group items by customer orders. On average., crisplant processes about 80% of the total outbound volume through its multipart operation flow. Because of high volume and complex process flow, the majority of defects, in RNO 1 FC. are seen in crisplant costing distribution center (RNO 1) significantly in labor hours. This research paper identifies and quantifies the major defects in crisplant, and outlines the solutions to reduce the cost of handling these defects in RNO 1. The project work thoroughly assesses the entire RNO 1 crisplant operations (induct, sort, pack, SLAM, and problem solve) through four-phase approach: Understand the crisplant Process Flow, Develop a Data Collection Framework, Collect and Analyze Data, and Identify/Implement Data Driven Solutions. Lean principles and methodologies were used throughout the project work especially when identifying solutions. For example, opportunities that improved the packing process were identified based on a deep-dive analysis as a part of the Kaizen study. The project results demonstrated 50% reduction in cost of handling crisplant defects in RNO l. Furthermore, it highlighted opportunities for additional savings by identifying solutions that can also be implemented in other FCs (i.e. SDF 1, TUL 1) with similar operation as RNO 1.