Chapter 18 Lean Operating Systems
COMPUTATIONAL PROBLEMS AND EXERCISES
These exercises require you to apply the formulas and methods described in the
chapter. The problems should be solved manually.
32. Esplanade Marine Parts, Inc. wants to produce parts in batches of 100
parts. Each part must be processed sequentially from workstation A to B.
a. How many seconds is required to produce the batch under the
assumptions of batch processing?
Workstation A is the bottleneck workstation. Here, the batch of 100 must be
completed at each workstation before it can be processed on the next
workstation or 35 seconds/part*100 parts = 3,500 seconds. Batching is useful if
changeover and setup costs are high.
b. How many seconds is required to produce the batch under the assumptions of
single-piece flow processing?
The first part completes Workstations A in 20 seconds and then moves to B,
finishing in 35 seconds. Since Workstation A is the bottleneck, parts cannot be
completed more than 20 seconds apart so parts 2 to 100 are processed in 20*99
= 1,980 seconds. Total is 2,015 (1980+35) seconds if single-piece flow. Notice
that workstation B is starved for work.
c. Compare the two solutions in terms of time saved and any other issue(s) you
think important.
Seconds (capacity) saved = 3,500 – 2,015 = 1,485. This is a 42.4% reduction in
total processing time (1,485/3,500 = 0.424). Many of the best practices shown
in Exhibit 18.6 are necessary to successfully implement one-piece flow.
33. CDC Discrete Fabricators wants to produce parts in batches of 300 parts.
Each part must be processed sequentially from workstation A to B to C to D.
a. How many seconds is required to produce the batch under the
assumptions of batch processing?
The bottleneck workstation is D. Here, the batch of 300 must be
completed at each workstation before it can be processed on the next
workstation or 70 seconds/part*300 parts = 21,000 seconds. Batching is useful
if changeover and setup costs are high.
b. How many seconds is required to produce the batch under the
assumptions of single-piece flow processing?
The first part completes Workstations A to C in 45 seconds and then
moves to D, finishing in 70 seconds. Since Workstation D is the bottleneck, parts
cannot be completed more than 25 seconds apart so parts 2 to 300 are
processed in 25*299 = 7,475 seconds. That is,
20 + 15 + 10 = 45 seconds for first three workstations; 25 seconds at
bottleneck workstation D
25*299 = 7,475 seconds for other parts (or since bottleneck is last
workstation some may multiply 300*25 = 7,500 + 45 seconds.
Total is 7,545 seconds if single-piece flow
c. Compare the two solutions in terms of time saved and any other issue(s)
you think important.
Seconds (capacity) saved = 21,000 – 7545 = 13,455. This is a 64%
reduction in total processing time (13,455/21,000 = 0.64). Many of the best
practices shown in Exhibit 18.6 are necessary to successfully implement one-
piece flow.
34. NOTE TO INSTRUCTORS. THIS PROBLEM HAS BEEN MODIFIED FROM THAT
IN THE ORIGINAL PRINT VERSION. IF YOU ASSIGN THIS, PLEASE ENSURE THAT
STUDENTS USE THE FOLLOWING.
A Florida lottery machine and printer at Store #8302 operates 360 days a
year. If the machine jams or breaks down, the Florida Lottery Commission loses
an average of $2,000 in revenue per day. Preventive maintenance can be
performed each morning and costs $670 per day, and the machine will not break
down. Without preventive maintenance, the probability of a breakdown is as
follows. What are the economics of the situation? What do you recommend?
The expected number of breakdowns per day without preventive
maintenance is 0*0.333 + 1*0.333 + 2*0.333 = 0.997. This means that the
machine will fail almost every day, resulting in lost revenue close to $720,000
annually. Since the Florida Lottery Commission loses nothing if preventative
maintenance is used, then the obvious decision would be to use preventive
maintenance at an annual cost of 360*$670 = $241,200.
35. A plastic injection molding machine operates 348 days a year. It is the
bottleneck workstation in a process that produces plastic housed wall clocks.
Industrial engineers collected data on the probability of machine breakdowns,
the costs of preventive maintenance, and the cost of a breakdown.
The cost of preventive maintenance is high due to performing it between 1
and 4 a.m. each day by two machine experts. The preventive maintenance cost
is $1,980 per day including replacement parts, cleaning, and software upgrades.
The cost of a breakdown of $1,850 is also high because the downstream
workstations can only work until all work-in-progress inventory is completed;
then the entire process must stop.
The expected number of breakdowns per day without preventive
maintenance is 1.31 (0*0.32 + 1*0.28 + 2*0.22 + 3*0.13 + 4*0.05). Preventive
maintenance cost is $1,980 per day. The expected cost of a breakdown per day
is $2,423.50 (1.31*$1,850). The cost of preventive maintenance is $1,980 per
day. The manufacturing firm saves $443.50 ($2,423.50 – $1,980) per day or
$154,336 per year by performing preventive maintenance daily on the injection
molding machine.
36. Bracket Manufacturing uses a Kanban system for a component. Daily
demand is 1000 units. Each container has a combined waiting and processing
time of .85 days. If the container size is 70 and the alpha value () is 13%, how
many kanban card sets should be authorized? What is the maximum authorized
inventory?
Using Equation 18.1 we have
K = d*(p + w)(1 + )/C = 1,000(0.85)(1 + 0.13)/70 = 13.7
(18.1)
Thus, 14 Kanban cards should be authorized. The maximum authorized
inventory is K*C = 14*70 = 980 units.
37. Tooltron Manufacturing uses a Kanban system for a component. Daily
demand is 425 units. Each container has a combined waiting and processing
time of 1.5 days. If the container size is 35 and the alpha value (a) is 12 percent,
how many kanban card sets should be authorized? What is the maximum
authorized inventory?
Using Equation 18.1 we have
K = d*(p + w)(1 + )/C = 425(1.5)(1 + 0.12)/35 = 20.4
Thus, 21 Kanban cards should be authorized.
The maximum authorized inventory is K*C = 21*35 = 1,116 units. You
might also use this simple problem to discuss “what if” scenarios regarding
increases and reductions in processing and wait times, impact of changes in
order size, safety factors, and so on. Exhibit 18.5 (a two-card Kanban
illustration) can be integrated into this discussion too.
38. Lou’s Bakery has established that JIT should be used for chocolate chips
due to the high probability of the kitchen heat melting the chips. The average
demand is 150 cups of chocolate chips per week. The average setup and
processing time is 1/4 day. Each container holds exactly two cups. The current
safety stock factor is 5 percent. The baker operates six days per week.
a. How many Kanbans are required for the bakery?
Using Equation 18.1 we have (note that setup and processing time is in
days and demand is in weeks):
K = d*(p + w)(1 + )/C = (150 cups/week/(6 days/week))(0.25 days)(1 +
0.05)/2 = 3.28 .
Therefore, 4 Kanbans should be used. Note: When using Equation 18.1 we
round to the next highest integer).
b. What is the maximum authorized inventory?
Maximum authorized inventory = K*C = 4*2 = 8 cups
c. If the average setup and processing time is increased to 3/8ths of a day
due to a process change, what are the answers to (a) and (b)?
K = d*(p + w)(1 + )/ C = (150 cups/week/(6 days/week))*(0.375
days)*(1 + .05) = 4.9, or 5 Kanbans
Maximum authorized inventory = K*C = 5*2 = 10 cups
By increasing p + w, inventory is increased from 8 to 10.
39. Due to rapid changes in technology, a telecommunications manufacturer
decides to produce a router using just-in-time methods. Daily demand for the
router is 17 units per day. The routers are built on racks that hold four at a time
(i.e., the container size). Total processing and waiting time is 3.75 days. The
process manager wants a safety factor of only 5 percent.
a. How many Kanbans are required?
K = d*(p + w)(1 + )/C = (17 units/day)(3.75 days)(1 + 0.05)/4 = 16.7
or 17 Kanbans
b. What is the maximum authorized router inventory?
Maximum authorized inventory = K*C = 17*4 = 68 units
c. If you assume that one-half of the racks are empty and one-half full
at any given time, what is the average inventory of routers?
Average inventory = maximum inventory divided by 2 = 68/2 = 34 units.
d. What are the answers to a to c if due to process improvements the
total processing and waiting time is reduced from 3.75 to 2.75 days?
K = d*(p + w)(1 + )/C = [(17 units/day)(2.75 days)(1 + 0.05)]/4 = 12.3
or about 13 Kanbans
Maximum authorized inventory = K*C = 13*4 = 52 units
Average inventory = maximum inventory divided by 2 = 52/2 = 26 units.
By reducing p + w, inventory is reduced from 34 to 26, or (34 – 26)/34=
42.6%
40. An automobile transmission manufacturer is considering using a JIT
approach to replenishing its stock of transmissions. Daily demand for
transmission #230 is 55 transmissions per day and they are built in groups of six
transmissions. Total assembly and waiting time is two days. The supervisor wants
to use an alpha value (α) of 1, or 100 percent.
a. How many Kanbans are required?
K = d*(p + w)(1+ )/C= [(55 trans/day)(2 days)(1 + 1)]/6 = 36.7, or 37
Kanbans
b. What is the maximum authorized inventory?
Maximum authorized inventory = K*C = 37*6 = 222 transmissions
c. What are the pros and cons of using such a high alpha value ()?
Pros - provides safety stock for emergency situations and process
breakdowns, keeps internal and external customers and suppliers happy, not
really JIT.
Cons - high inventory levels with increased number of Kanbans, may
require more debt financing of inventory, defeats the objectives of JIT systems,
clutters up shop floor and warehouse, higher risk of transmission obsolescence,
costly practice, and there is no numerical method to determine why it is set at
100%.
41. A pillow assembly line consists of five stages arranged in series. The first
workstation stuffs the pillow with filler material. The second workstation sews the
pillowcase and the third stage places the pillow in a clear plastic sleeve. The
fourth workstation places the pillow in a box of three for shipping and seals the
box. The fifth workstation places labels on the box for shipping. Pillows are low
value yet high cubic volume, so inventory holding cost is high. In fact, the
operations managers use a JIT lean assembly line and attempt to only produce
what they can sell. The assembly line works two shifts/day, 8 hours/shift minus 1
hour for breaks and food, and 330 days per year. Annual demand is 1,000,000
pillows. What is the Takt time in seconds? How is your answer used?
Takt time = Available time per time period/Market demand rate per time
period(18.2)
Takt time = [(7 hours/shift/day)(2 shifts/day)(330 days/year)(3600
seconds/hour)]/1,000,000 pillows sold/year = 16.632 seconds/pillow
Next the operations managers must make sure adequate resources are
available at each labor or machine constrained workstation so that each
workstation time is less than or equal to 16.632 seconds.
