An evaluation of routing and volume-based storage policies in an order picking operation

Decision Sciences, Spring 1999 by Petersen, Charles G II, Schmenner, Roger W

Across-aisle

In across-aisle storage with a corner PID, the highest volume item is stored in the first storage location of the first aisle. The next highest volume item is stored in the first storage location of the second aisle. Once the first storage location in all of the aisles is assigned an item, the second location of each aisle is assigned an item. The area close to the front aisle contains the high volume items and the area close to the back aisle contains the low volume items. With a middle PID point, the highest volume item is stored in the first storage location of the middle aisle. The next highest volume items are stored in the first storage locations in alternating order out from the middle aisle.

Perimeter

In perimeter volume-based storage the high volume items are located around the perimeter of the warehouse. The low volume items are placed within the middle of the aisles. The highest volume item is stored in the first storage location of the aisle closest to the PID point. The rest of the items are stored in the perimeter storage locations in a counterclockwise direction from the P/D point.

EXPERIMENTAL DESIGN

Table 1 presents the four factors and their associated factor levels used in the experiment. The routing heuristic factor (Route, 6 levels) includes composite (C), largest gap (LG), midpoint (M), return (R), transversal (T), and optimal (O). The storage policy factor (Store, 8 levels) consists of across-aisle corner P/D (AC), across-aisle middle PID (AM), diagonal corner P/D (DC), diagonal middle P/D (DM), perimeter corner P/D (PC), perimeter middle P/D (PM), within-aisle corner P/D (WC), and within-aisle middle P/D (WM). The pick list size factor (Pick, 5 levels) includes pick list sizes of 5, 15, 25, 35, and 45 items. Hall (1993) has shown that the number of picks has an effect on the performance of routing strategies. The demand skewness factor (Skew, 3 levels) consists of low, medium, and high demand skewness. The skewness of the demand for items is based on the realization that the demand for each item is not equal. Therefore, it is assumed that if items are ranked according to demand, that 20% of the items in the warehouse account for 40% (low), 60% (medium), or 80% (high) of the demand or pick activity in the warehouse.

There are 720 cells (6x8x5x3) and 30 replications per cell. However, each routing and storage factor level combination for a given pick list size and demand skewness factor level combination is tested on the same 30 randomly generated pick lists. This results in a mixed-model design, with routing and storage as the within-subjects factors and pick list size and demand skewness as the betweensubjects factors. The performance measure (route length in feet) is the total distance traveled by the picker to pick all items on the pick list. This distance includes the within-aisle distance within the front and back aisles and the across-aisle distance within the picking aisles.

RESULTS

The results of the experiment were analyzed by full factorial mixed-model ANOVA using SPSS for Windows (Release 6.1). The results are presented in Table 2. The analysis indicates that the main effects, route and store, exhibited a statistical significance of less than .01. The factors, pick and skew, are also significant; however, given the nature of these factors, these results are not surprising. In addition, all of the two-way and all but one of the three-way interactions are significant at an a of .01.