Inventory management
Background
Caterpillar Inc (CAT) is an American company that designs, develop, engineers, manufactures, markets , sells machinery and engines. It is first founded in 1925 when the Holt Manufacturing company and the C.L Best Tractor Company merge together and form California-based Caterpillar Tractor Company. Its headquarter locating in Deerfield, Illinois when prior to that, it was located in Peoria, also in illinois
As mentioned before, CAT is a company that is involved in numerous markets. It produces clothing, electronics, agriculture products, engine and gas turbines, trucks, defense products.
To expand its business scale, CAT is not afraid to do business in other developing markets. CAT built its first Russian facility in the town of Tosno, located near St. Petersburg Russia in 1999.
Another method that CAT often does is acquisition. From 1951 to 2020, CAT has acquired 39 companies where 2010 is the year where CAT is most active in acquiring companies. The companies are from numerous countries and continents. From this we can see the scale of CAT business is very wide-spread and is definitely a multi-national companies with well-established link and business partners throughout the globe.
Case study
In the lecture discussing about inventory management, we have encountered problems about inventory management of Caterpillar (CAT). CAT is a global producer of heavy machines such as tractors and loaders, reaching global sales of 53.8 billion US dollars in 2019.
CAT always strives to respond rapidly to customer request for spare parts and it always has to concern itself with 23 distributions in 11 countries.
In case the customer could not receive the spare part within
48 hours, the part has to be delivered free of charge as a customer service
guarantee. CAT needs to determine the
optimal values for when to order to restock the part in inventory and how much
to order when the order is placed.
In the case of the Belgium warehouse, they implement a continuous review reorder policy for the vast majority of its spare part inventory. In order to fulfill CAT's promises to send out spare parts within 48 hours, it needs to know 1. when to place an order to restock the part in inventory and 2. how many to order when the order is placed. Yet many of the parts are costly and troublesome. That includes a CAT 990 wheel loader with a capacity of 8m^3.
Parts of CAT 990 wheel loader
Holding cost=$135 per unit per year
Weekly demand =5 units per week in average with 3 units s.d.
Ordering cost=$1000 per order
Lead time =6 weeks
Backlog Penalty=$500/unit
Calculations
.png)
First, we need to find out the reorder level. Reorder point in unit can be calculated with lead time* demand rate, which is 20*10 =200 units
Uncertainty
Yet, demand can be unpredictable. We are taking the Belgium warehouse as an example.
|
EOQ |
Sqrt(2*1000*260/135) |
62 units per
order |
|
PR* |
135*62/500*260 |
0.0644 |
We
must put more variables into consideration. With the EOQ formula, we can substitute
the numbers into the formula to determine the EOQ. (2*1000*260/135)^(1/2)= 62
units /order.
Then we
need to determine the optimal reorder level R*. Step 1 of accomplishing that is
to calculate the probability first. We will time the holding cost with EOQ and
have it divided by backlog cost times annual demand. In the end we have 0.0644.
Plugging this number into ztable, we have z =1.52.
.png)
|
mean |
6*5 |
30 |
|
Standard deviation |
Sqrt(6)*3 |
7.35 |
Then we find the mean and sd of lead time demand. We do it by using the constant demand rate which is 5 units with 3 units of sd times lead time. 6*5=30. Standard deviation can be obtained by sqrt(6)*3=7.35.
|
R* |
30+(7.35*1.52) |
41 |
Then we can determine the reorder level R* by adding the mean to sd times the z value which is 41.
|
Q* |
Sqrt(2*(1000+500*0.21)*260/135) |
65 units per
order |
Then, we find
the standard normal loss value by plugging z value to the table and we have
0.02799. Then we translate that into nR=7.35*0.02799=0.21. We time nR with
backlog penalty and add it to fixed cost to place an order and plug it into the
formula sqrt(2*S*D/H).65 units order for Q*.
|
PR* |
135*65/500*260 |
0.0675 |
|
R* |
30+(7.35*1.49) |
41 |
.png)
With the
new sets of data, we can calculate the total inventory cost, C* is $10292.5.
ICE question:
1-liter can
|
EOQ |
sqrt( (2 * 7500 *
1000000) / 1 ) |
122474.4871 units |
|
Reorder |
1000000 / 12 |
83333.33333
units |
|
ttc |
(1 * 122474.4871) /
2 + 7500 * (1000000 / 122474.4871) |
$122474.49 |
5-liter drum
|
EOQ |
sqrt( (2 * 7500
* 200000 / 5) |
24494.89743 units |
|
Reorder |
200000 / 12 |
16666.66667
units |
|
ttc |
(5 * 24494.89743) / 2
+ 7500 * (200000 / 24494.89743) |
$122474.49 |
5-liter can
|
EOQ |
sqrt( (2 * 7500
* 300000) / 5 ) |
30000
units |
|
Reorder |
300000 / 12 |
25000
unit |
|
ttc |
(5 * 30000) / 2
+ 7500 * (300000 / 30000) |
$150,000 |
For the eoq for 1-litre can, it is 122474.4871 units with sqrt(2*(ordering cost)*(average demand)/(holding cost)). The reorder point of 1-liter can is demand/12 months 1000000/12=83333.33. The TTC is 122474.49. For the eoq for 5-litre drums: sqrt(2*7500*200000/5)=24494.8973 units. The reorder point for it is 16666.6667units and the TTC is 122474.49. EOQ for 5-liter cans is 30000 units and the reorder point is 25000units and the TTC is 150000.
Q.2
|
1-litre can |
Pr =
(1 * 122474.4871) / (2 * 1000000) = 0.0612 |
z = 1.545 |
|
5-litre
drums |
Pr =
(5 * 24494.89743) / (10 * 200000)= 0.0612 |
z = 1.545 |
|
5-litre
cans |
Pr = (5 * 30000)
/ (10 * 300000) = 0.05 |
z = 1.645 |
Under uncertain demand, we must calculate the probability that the demand is larger than the reorder point, Pr for 1-liter can be holding cost *eoq divided by stock out unit cost times average annual demand= 1.545,5-litre drum is 1.545 and 5-liter cans is 1.645.
1-litre can
|
Mean |
1,000,000/12 |
83,333 |
|
SD |
15,000/sqrt(12) |
4330 |
|
Mean |
200,000/12 |
16,667 |
|
SD |
5,000/sqrt(12) |
1,443 |
|
Mean |
300,000/12 |
25,000 |
|
SD |
10,000/sqrt(12) |
4330 |
|
1-litre
can |
83,333+4,330*1.54 |
90,000 units |
|
5-litre
drums |
16,667+1443*1.54 |
18,890 units |
|
5-litre
cans |
25,000+2,887*1.64 |
29,735 units |
We previously made the mistake of using the S.D provided yet we forgot to divide it by the square root of 12 first. Therefore, we answered the question incorrectly.
The reorder level is 90,000, 18,890, and 29,735 respectively.
Then using the z value we obtained, we translate the z value into standardized normal loss value and times them with S.D.
|
1-litre
can |
Sqrt((2*(7500+2*115.7)*1,000,000)/1) |
124,350 units/
order |
|
5-litre
drums |
Sqrt((2*(7500+10*38.6)*200,000)/5) |
25,117units/order |
|
5-litre
cans |
Sqrt((2*(7500+10*61)*300,000)/5) |
31,196 units/order |
With the new information, we can re-evaluate Q*.
|
1-litre
can |
1*124,350/2*1,000,000 |
0.06217 |
|
5-litre
drums |
5*25,117/10*200,000 |
0.06279 |
|
5-litre
cans |
5*31,196/10*300,000 |
0.052 |
Then we calculate the probability
|
1-litre
can |
Same value |
Same value |
|
5-litre
drums |
16,667+1443*1.53 |
18,875 units |
|
5-litre
cans |
25,000+2,887*1.63 |
29,705 units |
new R*
|
1-litre
can |
131,017 |
|
5-litre
drums |
136,702 |
|
5-litre
cans |
179,656 |
With the new data, we can calculate the new ttc
Q.3 this policy collects all different classes of oil into one class. The chance of running out of specific types of classes of oil is eliminated. This gives us some insight as to how to reduce the chance of running out of a specific type of inventory is to simplifying specific types of inventory and classifying them under one category.
Reflection
From this lecture content, we can see the inventory management system has become way more costly when we add in some additional factor into consideration.
First, when we look at the eoq without considering the probability of CAT running out of stock and we have the answer 62 units per order. However, when we consider the probability that we may run out of stock and add the backlog penalty times the probability to it and the unit has increased by 3 more units to dampen the effect of uncertainty demand. The cost is not only ordering 3 more units but also the inventory cost that CAT needs to hold those stocks inside its warehouse including but not limited to maintenance cost. Also, if the demand for some reason declines, CAT would have abundant stock left in the warehouse and they may need additional spaces to house the excessive components.
Second, we will need to consider the probability where the demand is larger than supply. Previously, we have 0.0644 as the probability but after we times holding cost with the renewed Q, the probability of running out of stock becomes 0.003 higher approximately, which is 5% higher chance to run out of stock. It may be a small sum to CAT. Yet it could be detrimental to its reputation and branding.
Through this lesson do we know that the importance to analyze data at hand to try to minimize the risk of failures as it could not only cause monetary losses as a company would need to compensate for the failure to deliver to its customers and lose its reputation.
With this in mind, CAT seems to be actively continuing to refine its process to delivery the best service to its customers.
Additional information
Caterpillar and Ford are joining hands to develop a new management system that enables more efficient and swift delivery of spare parts to their customers called SAP. With this new system in mind, the travel time in Caterpillar itself will be drastically decreased. Therefore, the need to hold safety stock can be drastically decreased as the travel speed is greatly enhanced. Possibly, the inventory holding cost can be reduced because with safety stock being decreased, the duration of an item having to stay in the warehouse is drastically decreased. Therefore, the resources spent on maintaining the parts in the warehouse is decreased as well.
Advantages
This gives a huge edge to Caterpillar over its competitors. According to Boston-based AMR Research Inc. analyst Kevin Prouty, the other companies that can create a system similar to SAP are IBM corp or EDS( Electronic Data System). Yet SAP can offer more of a packaged application than either of those two companies can offer. With a similar system so hard to obtain, Caterpillar competitors, mainly Komatsu and Volvo, without the system, would have to spend more on maintaining its inventory, and keeping more safety stock due to the longer lead time which brings more uncertainty when it comes to delivery. It not only brings the cost higher when compare to caterpillar but also their reputation as Caterpillar new system can allow customers to track the whereabouts of the spare parts which instills more confidence to its customers.
-----------------------------------------------------------------------------------------------------------------------------
References
https://www.caterpillar.com/
https://www.computerworld.com/article/2577634/caterpillar--ford-link-up-for-spare-parts-management-system.html
https://businesschronicler.com/competitors/caterpillar-competitors-analysis/#:~:text=Answer%3A%20Caterpillar's%20main%20competitors%20are,%2DInternational%20AG%2C%20and%20SDLG.
.svg.png){kind=logo}
留言
發佈留言