SCMA 6330 - Business Logistics Systems

LOGISTICS ANALYSIS - TUCKERMAN DISTRIBUTING

Tuckerman Distributing is regional distributer of a wide variety of supplies used in residential and commercial construction in the eastern US. They are a well-established company with customers in many cities, but they are facing strong competition from newer distributors, especially those using online sales and marketing. However, Tuckerman Distributing expects their business to grow as long as they can be effective in maintaining low costs and improving service through competitive delivery times. Tuckerman's current distribution system uses a single large distribution center (DC) in Chicago. This DC was an outgrowth of the founding of the company (as a local distributor in Chicago), and it has been retained in the belief that a single large DC provided the best opportunities to maintain administrative control and exploit economies of scale to keep costs low. However, in light of the recent strong increase in competition and the importance of service (reducing delivery times) Tuckerman is seeking a redesign of their distribution system. As a pilot test, they wish to consider how best to serve a subset of their customers consisting of 24 important markets (cities), mainly in the northeastern US. The 24 cities to be served are shown in Table 1 along with the annual demand from each city. Demand is measured in terms of average orders. (Individual orders vary from a few pounds to many hundreds or thousands of pounds, so the demand in Table 1 is in terms of the number of average orders per year for each city.)

Tuckerman in considering opening one or more new DCs and has identified six possible locations for these. These six locations along with the Chicago DC are shown in Table 2. This table includes the supply that would be available from each location based on the projected size and the annual fixed cost of the DC.

Tuckerman uses a variety of trucking firms to transport orders from the Chicago DC to their customers and they plan to continue this arrangement of truck transportation with the new DC(s). They have assembled an estimate of the transport cost for the current and potential new DCs ($ per order to ship from any DC to the customers) and it is shown in Table 3. However, with the fluctuations in fuel prices, the transport costs may actually differ from those in Table 3. This variation in transport costs should be considered in your analyses.

In designing their new distribution system Tuckerman is naturally concerned with keeping costs low, and the data provided includes annual fixed facility costs and transportation costs (Tables 2 and 3). But Tuckerman is increasingly concerned about the level of service as competitors appear to be planning to offer high service levels (mostly 1 day delivery) to these cities in the near future. To help assess the service levels, Tuckerman has compiled truck travel time information for the DCs and cities being served in Table 4. This shows that Chicago is two days away from many cities, and even three days from a few. Tuckerman is not sure how best to measure the level of service. One simple idea is to use the average delivery time (a weighted average based on the orders being sent). But Tuckman is also interested in a more refined approach based on the % of orders (or revenue from customers) delivered in 1 day, 2 days, and 3 days, especially because in the near future the standard level of service may be "1 day".

Table 1: Stores and Demand

Cities	Demand (orders per year)
Albany, NY	4700
Allentown, PA	2400
Baltimore, MD	9200
Barnstable, MA	3000
Boston, MA	64400
Buffalo, NY	20900
Burlington, VT	4600
Chicago, IL	46700
Columbus, OH	9500
Erie, PA	2800
Harrisburg, PA	2200
Hartford, CT	9200
Lancaster, PA	2600
Lewiston, ME	2000
London, KY	6200
Manchester, NH	3600
New Haven, CT	5600
New York, NY	49500
Philadelphia, PA	28700
Pittsburgh, PA	17400
Portland, ME	3300
Providence, RI	11200
Rochester, NY	8300
Washington, DC	22300

Table 2: Possible DC Locations, Supply (in orders) and fixed costs ($/year)

	Allentown, PA	Chicago, IL	Bridgeport, CT	Baltimore, MD	Buffalo, NY	Andover, MA	Pittsburgh, PA
Supply	100,000	400,000	100,000	100,000	100,000	100,000	80,000
Fixed Cost	500,000	1,000,000	400,000	500,000	500,000	400,000	250,000

Table 3: Shipping Cost ($ per Order)

	Allentown	Chicago	Bridgeport	Baltimore	Buffalo	Andover	Pittsburgh
Albany, NY	11.2	13.3	11.2	11.5	11.5	11.2	12.6
Allentown, PA	9	11.1	9	9	9.3	9.3	9.3
Baltimore, MD	9	10.4	9.3	9	9.3	10.4	9.3
Barnstable, MA	9.3	11.1	9	10.4	10.4	9	10.4
Boston, MA	9.3	11.1	9	10.4	10.4	9	10.4
Buffalo, NY	11.5	12.6	11.5	11.5	11.2	12.6	11.5
Burlington, VT	9.3	11.1	9.3	10.4	9.3	9	10.4
Chicago, IL	11.1	9	11.1	10.4	10.4	11.1	10.4
Columbus, OH	12.6	11.5	12.6	12.6	11.5	13.3	11.2
Erie, PA	11.5	11.5	12.6	11.5	11.2	12.6	11.2
Harrisburg, PA	11.2	12.6	11.5	11.2	11.5	11.5	11.5
Hartford, CT	9	12.6	9	9.3	10.4	9	10.4
Lancaster, PA	9	11.1	9	9	9.3	9.3	9.3
Lewiston, ME	10.4	11.1	9.3	10.4	10.4	9	10.4
London, KY	13.6	11.8	14.3	13.6	13.9	13.3	11.7
Manchester, NH	9.3	11.1	9	10.4	10.4	9	10.4
New Haven, CT	9	11.1	9	9.3	9.3	9	10.4
New York, NY	11.2	13.3	11.2	11.2	11.5	11.5	11.5
Philadelphia, PA	11.2	13.3	11.2	11.2	11.5	11.5	11.5
Pittsburgh, PA	11.5	12.6	12.6	11.5	11.2	12.6	11.2
Portland, ME	12.6	13.3	11.5	12.6	12.6	11.2	12.6
Providence, RI	9.3	11.1	9	10.4	10.4	9	10.4
Rochester, NY	11.2	12.6	11.5	11.5	11.2	11.5	11.5
Washington, DC	11.8	10.4	11.8	11.8	11.1	11.8	11.1

Table 4: Delivery Times (1 = next day, 2 = 2 days, etc.)

	Allentown	Chicago	Bridgeport	Baltimore	Buffalo	Andover	Pittsburgh
Albany, NY	1	2	1	2	1	1	2
Allentown, PA	1	2	1	1	2	1	2
Baltimore, MD	2	2	2	1	2	1	1
Barnstable, MA	1	3	1	2	2	2	2
Boston, MA	1	3	1	2	2	2	2
Buffalo, NY	2	2	2	2	1	1	1
Burlington, VT	1	3	1	2	2	2	2
Chicago, IL	2	1	2	2	2	2	2
Columbus, OH	2	2	2	2	2	2	1
Erie, PA	2	2	2	2	1	2	1
Harrisburg, PA	2	2	2	1	2	1	1
Hartford, CT	1	2	1	2	2	1	2
Lancaster, PA	2	2	1	1	2	1	1
Lewiston, ME	2	3	1	2	2	2	2
London, KY	2	2	2	2	2	2	2
Manchester, NH	1	3	1	2	2	2	2
New Haven, CT	1	2	1	2	2	1	2
New York, NY	2	2	1	1	1	1	1
Philadelphia, PA	2	2	2	1	2	1	1
Pittsburgh, PA	2	2	2	2	2	2	1
Portland, ME	1	3	1	2	2	2	2
Providence, RI	1	3	1	2	2	2	2
Rochester, NY	2	2	2	2	1	1	1
Washington, DC	2	2	2	1	2	1	1

Tuckerman knows that to stay in business they must offer a higher level of service. You are a consultant hired by Tuckerman Distributingto make recommendations regarding their distribution system. Address the questions below using discrete facility location models developed in Excel. For each question, describe the preferred plan (locations of DCs, shipping patterns, costs, service, etc.) and explain why this is your recommendation.

In solving your models with the basic version of Solver and Excel, you will need to make a few changes. In Solver Options, under Constraint Precision:

Use Automatic scaling (check the box)

Set Integer Optimality (%) to 0.001

In Solver Options, under Constraint Precision:

Set Max Time (Seconds) to 240

Set Iterations to 10000

For some large or difficult problems you may need to expand these values.

1. First, as a baseline, develop a discrete facility location model in Excel and solve it to determine the minimum total cost corresponding to the current situation of using only the Chicago DC. (To verify your model is correct, this should be a total cost of $4,943,110/year - and the average delivery time for this solution is 2.133 days.). Then address the questions below that allow new DCs. Consider performance measures for cost (total cost and average cost per order) and service.

a. If any of the DCs can be used, which DCs should be open and what are the shipping patterns? What is the cost and the level of service? How does this compare to the solution with Chicago DC only in terms of cost and service?

b. Suppose Tuckman wants to consider closing the Chicago facility and serving these 24 cities from the new DCs. Which new DCs should be open and what are the shipping patterns? What is the cost and the level of service? How does this compare to the solution with the Chicago DC only and with #1a in terms of cost and service?

c. Suppose Tuckman wants to consider a more balanced network with all seven DCs open where each must serve at least 40,000 orders. What are the shipping patterns? What is the level of service? How does this compare to the solutions above.

2. To improve service Tuckerman believes new DCs should be opened. So consider the bestservice (minimize average number of days for deliveries) solutions with 1, 2, 3, 4, 5, 6 or 7 DCs open. (This does include the condition that each DC serve at least 40,000 orders.)

a. Make a chart that plots cost and service using two axes for these seven best service solutions with 1-7 DCs. Also, using the same data, make a chart that plots one line showing cost (vertical axis) vs service (horizontal axis) for the seven solutions.

b. Based on these seven results, how many DCs seem best? Which DCs should be open, what are the shipping patterns, and what is the cost and the level of service?

3. Tuckerman would like to better understand the relationship between cost and service. They expect that higher service networks will increase the costs, but they do not know how much; and they are uncertain how to measure service. Service could be measured as average days for a delivery, or in terms of the number or percentage of orders delivered in 1, 2 or 3 days, or perhaps some other ways. Perform a more refined analysis of cost and service using the facility location model to solve a number of scenarios with varying service levels, or varying costs, to map out in more detail the cost and service tradeoffs. Report the cost and level of service for each configuration.

a. Produce a chart that shows the tradeoff between cost and service.

b. From the various solutions that you find, what would you recommend as the best network for Tuckerman?

4. Some executives at Tuckerman think that to remain competitive, the percentage of orders delivered in one day should be at a very high level. So for this question, consider the highest possible level of service to maximize the percentage of orders delivered in 1-day. In comparison to the current situation where all customers are served from Chicago, this highest service level network will result in some (but not all) customers experiencing faster delivery. Tuckerman believes that offering such a high level of service could result in an increase in demand from the customers whose service improves, and they wonder what network design would be best to handle such an increase in demand. Suppose that customers that move from 2-day delivery currently (with only the Chicago DC) to 1-day delivery (with the highest service level network) will generate a 50% increase in demand over 3 years. Similarly, customers that move from 3-day delivery currently (with only the Chicago DC) to 1-day delivery (with the highest service level network) will generate a 100% increase in demand over 3 years. (For example, if Albany NY receives 2-day delivery currently, but moves to 1-day delivery with the high service level network, then the Albany demand after 3 years would be 47001.5 = 7050.)

a. Perform an analysis of the situation with the increased demand from the higher level of service.

b. With this possible demand increase from the higher level of service, what do you recommend?

5. In comparing the various recommended distribution systems from the questions above, what is your overall recommendation for Tuckerman's distribution network? Describe the preferred plan (locations of DCs, shipping patterns, etc.) and explain why this is your recommendation.

6. What additional key issues should be considered in analyzing Tuckerman's distribution problems?