A case study on order picking schedule deviations and their contributing factors

Abstract

Efficiency in order picking is crucial amid rising competition and customer expectations. It is common practice in warehouse literature to portray the order picking process as deterministic and fully predictable. However, this assumption is in most cases inconsistent with reality, leading to inaccurate system modelling. This case study analyses real-life data to identify and explore drivers behind deviations from the predetermined process flow in a manual order picking context. A generic methodology to learn deviations in both picking order and execution time is proposed, providing a framework for analogous data sets within diverse organisational contexts. Applied to a real-life data set of about three million picks performed by over 500 order pickers, the results indicate that several task- and human-related factors may drive deviations, affecting output predictability and overall efficiency. This case study empirically demonstrates that stochasticity is a significant yet often underestimated characteristic of order picking systems, as order pickers may ignore routing guidelines and may have varying working paces. The study concludes that data-driven decision-making can identify and help understand process deviations, leading to improved efficiency, cost savings, and potentially higher worker satisfaction by aligning managerial expectations with real-life performance.