Random vs Class-based allocation policies: impact of the warehouse parameters on the distance traveled by pickers

Within most distribution centers, activities are managed with low levels of automation, often becoming cost and labor-expensive due to the occurrence of inefficiencies. Picking and product allocation are the activities most likely to be subjected to these risks and, in literature, numerous contributions aim to optimize them. The present study aimed to evaluate the impact that different strategic variables, such as the routing policy (RP), the warehouse shape factor, the numerosity of the picking list, or the location of the Input/Output (I/O) points, have on the variability of the average distances traveled by pickers to fulfill a picking mission (d). This evaluation was done by comparing, through a simulation approach, different levels of class-based allocation policies versus a random allocation policy under different operating scenarios. An ANOVA was then performed to rank the variables and their interactions according to their impact on d in each scenario. The RP turned out to be the parameter with the strongest impact, covering between 23% and 36% of the variance; at the same time, moving from mild to strong class-based scenarios, the impact of the interaction of I/O and RP increased more than that of I/O alone. Possible future research activities were finally outlined.