.png){kind=icon}
.png){kind=icon}
Return
Optimizing workforce allocation under uncertain activity duration
.png){kind=icon}
OA
AI
DOI：10.1016/j.cie.2023.109228.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Even though warehouses are becoming increasingly automated, humans remain their central and most important resource. Every day, various activities must be carried out by workers. The assignment of individual workers to specific tasks has a major impact on the overall efficiency of a warehouse. The problem of finding an efficient assignment is not trivial and is complicated by task durations being unknown in advance, operational constraints, and the fact that employee well-being must be taken into consideration to maintain employee satisfaction. The method proposed in this paper uses work profiles: ordered lists of task properties such as type and work zone. Each worker is assigned exactly one profile and tasks are dynamically allocated to workers based on their profile. Finding a good profile assignment is crucial, yet the profiles are usually assigned manually by shift supervisors. This paper proposes a framework for automating the assignment of profiles to employees under uncertain task durations. The proposed approach applies a metaheuristic algorithm together with discrete-event simulation to evaluate the quality of a solution. The simulation component is used to address uncertainty of the task durations either by considering multiple scenarios or by using mean task durations. Possible values for task durations originate from distributions which are assumed to be given or learned from historical data. The contributions of this paper are threefold: (1) we propose a profile-assignment framework that deals with uncertainty of the task durations, (2) we study the trade-offs between run time and accuracy within this framework, and (3) we analyze our main design decisions and demonstrate how our method outperforms reconstructed solutions produced by a human expert.
Keywords:
Warehouse scheduling
Uncertainty
Heuristics
Simulation
Decision tree
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.