This research investigates the warehousing operations of internet retailers. The primary physical process in internet retail is fulfillment, which typically involves a large internet fulfillment warehouse (IFW) that has been built and designed exclusively for online sales and an accompanying parcel delivery network. Based on observational studies of IFW operations at a leading internet retailer, the investigations find that traditional warehousing methods are being replaced by new methods which better leverage information technology and efficiently serve the new internet retail driven supply chain economy. Traditional methods assume a warehouse moves bulk volumes to retail points where the bulks get broken down into individual items and sold. But in internet retail all the middle elements of a supply chain are combined into the IFW. Specifically, six key structural differentiations between traditional and IFW operations are identified: (i) explosive storage policy (ii) very large number of beehive storage locations (iii) bins with commingled SKUs (iv) immediate order fulfillment (v) short picking routes with single unit picks and (vi) high transaction volumes with total digital control. In combination, these have the effect of organizing the entire IFW warehouse like a forward picking area. Several models to describe and control IFW operations are developed and optimized. For IFWs the primary performance metric is order fulfillment time, the interval between order receipt and shipment, with a target of less than four hours to allow for same day shipment. Central to achieving this objective is an explosive storage policy which is defined as: An incoming bulk SKU is exploded into E storage lots such that no lot contains more than 10% of the received quantity, the lots are then stored in E locations anywhere in the warehouse without preset restrictions. The explosion ratio Ψo is introduced that measures the dispersion density, and show that in a randomized storage warehouse Ψo <0.01, whereas in an IFW the likely range is Ψo>0.40.
Specific research objectives that are accomplished: (i) Develope a descriptive and prescriptive model for the control of IFW product flows identifying control variables and parameters and their relationship to the fulfillment time performance objective, (ii) Use a simulation analysis and baseline or greedy storage and picking algorithms to confirm that fulfillment time is a convex function of E and sensitive to K, the pick list size. For an experimental problem the fulfillment time decrease by 7% and 16% for explosion ratios ranging between Ψo=0.1 and 0.8, confirming the benefits of an explosive strategy, (iii) Develope the Bin Weighted Order Fillability (BWOF) heuristic, a fast order picking algorithm which estimates the number of pending orders than can be filled from a specific bin location. For small problems (120 orders) the BWOF performes well against an optimal assignment. For 45 test problems the BWOF matches the optimal in 28 cases and within 10% in five cases. For the large simulation experimental problems the BWOF heuristic further reduces fulfillment time by 18% for K =13, 27% for K =15 and 39% for K =17. The best fulfillment times are achieved at Ψo=0.5, allowing for additional benefits from faster storage times and reduced storage costs.
If you have any questions please contact the ETD Team, libetd@njit.edu.
![[NJIT logo]](http://archives.njit.edu/vhlib/images/njit_logo.gif){kind=logo}
![[Van Houten Library logo]](images/vhl-top-image.jpg)
