The transshipment model involves the addition of an intermediate node to the path of a shipment. Nodes are usually added in order to increase flexibility and add a dynamic layer to shipment design. By incorporating nodes, shipments can be consolidated or split up at the intermediate destination in order to better accommodate demand point or final destination requirements. Operations managers or logistics planners may further regulate or better control flow based on strategic drivers such as supply chain disruptions, commodity shortages, weather or even regional disasters.
If we re-visit my previous post RE: the expanding motorsports parts company in New Mexico, we can apply the transshipment model to optimize their inbound parts shipments. Bartz Motorsports Parts and Accessories (BMPA) has two retail sites; one in Las Cruces, NM and an expansion retail site in El Paso, TX.
Scenario:
Over the past six months, BMPA has been unable to stock a sufficient supply of its fast moving parts and accessories. This is likely due to Supply chain disruptions brought about by the global pandemic and the poor national response to the supply chain disruptions. In an effort to increase throughput and overcome supplier challenges, BMPA seeks out and successfully contracts a secondary parts supplier in Houston, TX. The secondary supplier in Houston has the capability to successfully augment BPMA’s primary supplier in Dallas, TX and thus alleviate the gap in part availability and depth of stock.
Meanwhile in El Paso, BMPA’s retail expansion site has exceeded expectations. The West Texas market for specialized motorsports parts is booming as a growing semi-professional motocross competition circuit has driven demand to unprecedented levels. Sales for the El Paso retail site nearly matched the main store in Las Cruces where BMPA dominates the regional market. It’s a good problem to have, but BMPA is now concerned that it may not be able to ensure the 24 hour delivery of its special order kits, lean jet kits and carbon fiber brake systems. Or can it? We know that BMPA already employs best practices in inventory management by managing an authorized stockage list that factors in highly variable demand parameters. Other best practices ensure that it meets demand under normal conditions, but what about during the current supply chain disruption?
The addition of the secondary supplier in Houston will pave the way for the depth of stock needed to support BMPA’s two booming retail stores. Working with a 3PL, BMPA sets up two transshipment points or nodes in the city of San Antonio that would be a node for the Houston supply lane and another in Midland-Odessa for the Dallas supply lane. These nodes are especially useful in Texas where vast distances of travel are required between East and West Texas. By segmenting the journey, additional opportunities are provided for Trans loading, splitting and consolidating shipments; flexible transportation is always a preferred methodology. Most importantly, the transshipment points empower BMPA planners through their 3PL to completely regulate inbound parts shipments. They have chosen adequate warehouse space in San Antonio and Midland-Odessa that provides storage capability for up to 50 pallet positions (48” x 40”) at each location. Now, BMPA will have added storage capacity for the fast moving lean jet kits, carbon fiber brake systems and any other highly specialized parts that it deems necessary to retain in depth. Furthermore, the two node warehouse annexes are closer to their respective retail stores and located in more affordable areas of Texas.
BMPA benefits the most from 1) fully regulating inbound parts flow and 2) the newly acquired emergency reserve parts (uncommitted). As previously mentioned, this allows BMPA to regain its lost power due to unforeseen circumstances such as the disruptions brought about by the pandemic. These countermeasures may also protect product flow during weather events, especially as Dallas and much of East Texas resides in tornado alley. The flow plan should be reviewed and revised at least monthly in order to best utilize the transshipment model. The transshipment problem should be solved along with each monthly review and analysis.
There are many ways to solve the transshipment problem, where the supply and demand is balanced overall at 700 units/week, the distribution by retail site is not balanced to align with the corresponding supply point. In this example, the assignments are less relevant as all shipments have to be routed to their retail sites along the shared interstate 10 through El Paso, geography matters. The primary purpose of balancing the distribution is to meet the varying demands of the retail sites, but another important reason is to control the costs of shipping and maintain efficiency throughout the transshipment sequence.
Admin. note: the best way to solve transshipment problems is simply in excel using the “solver” add-in. If you don’t already have this add-in, it just takes a minute to install, simply follow this file path in excel: File – Options – Add Ins – Excel Add-Ins – Solver – Go. As an aside, data scientists and people living in academia also solve transshipment problems and they love to use the high performing MATLAB language and proprietary computational software. For transshipment problems, using MATLAB is kind of like flying a supersonic jet from Los Angeles to San Bernardino to save time…it’s a bit much, very impressive platform.
Build a diagram showing the shipment flow for context and then build a work area for the arcs and the nodes to break up the work a bit. First compute the decision for how much to ship for each shipping lane in the arc work area. I only added the primaries for this example; suppliers to final destinations. In D22, the objective function cell use formula “=SUMPRODUCT (Ship, Cost) with named ranges instead of cell references for ease. Enter your values and determine your total transportation cost. Next move over to the nodes, which are meant to represent what is coming to the node and what is leaving the node, again, I abbreviated the example here. This is where the balancing is computed and solved for us. The inflow and outflow can be computed using the =SUMIF(To,Node Reference,Ship) function in excel. Specifically, this would be =SUMIF(To,G17,Ship)-SUMIF(From,G17,Ship). In excel solver, set your objective as the total transportation cost (D22) – make sure you set to “minimize; min. Next enter your changing variable cells as your shipping range, that’s A17 – A20 or $A$17:$A$20 specifically. Lastly, add your constraints, in the drop down which are = to the supply demand as they are balanced. The constraint entry should look like this: $G$17 = $G$18. Finally, check the box for “make unconstrained variables non-negative and your solving method should be “simplex LP”, LP = linear programming. Solver will populate your A17-A22, G17, and G18 and provide a total transportation cost. The answer or A17-A20 will tell you how many units to ship from each supply point.
I sketched some broad strokes with this simple example but you can get an idea of how to optimize using transshipments and stay balanced and optimized as conditions change using solver. Excel is a very powerful tool with the solver and the data analysis tool pack that seem to be underutilized. The play by play explanations in a blog post are not as helpful as a video explanation in Adobe Presenter. I will put a video together and talk this through in much greater detail in a future post. It may be interesting to solve a transportation problem, then expand the problem by adding nodes and solve a transshipment problem in the same presentation. Reach out for corrections, questions or general commentary.