Dynamic Workforce Scheduling and Relocation in Hyperconnected Parcel Logistic Hubs

TL;DR

This paper introduces a dynamic, sensor-driven workforce scheduling and relocation system for parcel logistics hubs, improving flexibility and efficiency amid demand uncertainties during e-commerce surges.

Contribution

It presents novel reactive heuristics for real-time workforce and location optimization in hyperconnected logistics hubs, leveraging updated data and demand predictions.

Findings

Significantly outperforms traditional scheduling methods.

Reduces scheduling costs and improves parcel sorting timeliness.

Demonstrated effectiveness on real-world Chinese logistics networks.

Abstract

With the development of e-commerce during the Covid-19 pandemic, one of the major challenges for many parcel logistics companies is to design reliable and flexible scheduling algorithms to meet uncertainties of parcel arrivals as well as manpower supplies in logistic hubs, especially for those depending on workforce greatly. Currently, most labor scheduling is periodic and limited to single facility, thus the number of required workers in each hub is constrained to meet the peak demand with high variance. We approach this challenge, recognizing that not only workforce schedules but also working locations could be dynamically optimized by developing a dynamic workforce scheduling and relocation system, fed from updated data with sensors and dynamically updated hub arrival demand predictions. In this paper, we propose novel reactive scheduling heuristics to dynamically match predicted arrivals with shifts at hyperconnected parcel logistics hubs. Dynamic scheduling and allocation mechanisms are carried out dynamically during delivery periods to spatiotemporally adjust the available workforce. We also include penalty costs to keep parcels sorted in time and scheduling adjustments are made in advance to allow sufficient time for crew planning. To assess the proposed methods, we conduct comprehensive case studies based on real-world parcel logistic networks of a logistic company in China. The results show that our proposed approach can significantly outperform traditional workforce scheduling strategies in hubs with limited computation time.