Lean Production

Week 5: Production smoothing

Tran Van Ly
Industrial Engineering & Management
International University
Email: [email protected]
21/04/08 Room: A2-504 1
 Recall previous week
Cycle time, Flow time, Setup/Changeover time
1.Determine takt time

2.Determine completion time per unit

3.Determine standard operation routine

4.Determine standard quantity of WIP

5.Prepare the standard operation sheet

Processing time & Material handling

2
 Recall previous week
 Calculate the following items:
 Required run time/day: part A: = 250x40/60
=167min, part B=263min
 Time required to meet demand: 167+263 =
430min
 Available time for changeovers: 7.5x60x2 –
430-40 = 430 min
 Number of changeover per day = 430/50 =8.6 =
8 or 9
 (Calculating available time for changeover time
for individual product, it may underestimate)
3
 Learning Objectives
Methods to minimize quantity variance in a
production line
Leveling

Batch Sizing

Sequencing

4
 Why production smoothing?
Production smoothing: Minimize quantity variance
in a production line
Most important condition for JIT production
Production is promptly adaptable to demand
changes

 1. Smoothing of the total production quantity

 2. Smoothing of every model’s production

quantity
 1. Smoothing of Total Production
Quantity
 To produce the same amount of products
every period
 Allow daily production volumes to remain
constant despite demand varying (level the
total output in a month)
 Have per month quantitative production
schedule based on demand forecast 
divided by the number of operating days in
the month to get a daily production volume
 Ex: Master production plan
divided into two-week intervals
Date 1 2 3 … 14 15 16 17 18 . 30 31

Production 250 245 245 … 250 250 205 200 200 . 205 205
quantity
Total no. of 54 54 54 54 54 44 44 44 . 44 44
workers
Cycle time 120 120 120 .. 121 121 140 144 144 . 140 140
 What-if
If the quantity of demand within a month is not
constant?
The shorter the period of a master production plan,
the better for executing smoothing of the total
production quantity
However, time span of the master production plan
also cannot be too short, else smoothing of the total
production quantity will disappear
 Adapting to increased demand
 Use early attendance & overtime
 Improvements within each process also produce extra
time
 Adapt operation according to monthly schedule for
average daily production
Example:
Load on machine ordinarily is set 90% of its full capacity,
each multifunctional worker handle 10 machines
When demand increases, temporary workers are hired,
each handles less than 10 machine, utilizing 100%
machine capacity
 Adapting to decreased demand
• In parts manufacturing processes, the number of machines
handled by each worker increase
• On the assembly line, cycle time will increase due to reduce
demand quantity
• Developing standardized work in advance for different takt
scenarios
• Let extra workers involve other activities than to produce
unnecessary stock
 Activities may be organized
during a slack period
 Transfer workers to other lines for which demand
increases
 Decrease overtime
 Use a paid holiday
 Conduct quality control circle meetings
 Practice set-up actions
 Conduct maintenance and repair of machines
 Manufacture improved tools and instruments
 Conduct plant maintenance and upkeep
 Manufacture pars previously purchased form
suppliers
 2. Smoothing each model’s
production quantity
What would happen if the final assembly line
produced one type of product all day long?
 Volume: Takt time
 Style?
 Example - Leveling
 1824 units/month; 16 work days/month
 One shift per day (6am-4:30pm) 30
minutes lunch, 2x15 minutes break
 1824/16 = 114 units/work day
 10.5*60-60 = 570 minutes/work day
 Takt time = 570/114 = 5 minutes
 If there are three styles A, B and C (56,
29 & 29 respectively per day)
 If build in complete batches

 Production Sequence:
A1 to A56 then B1 to B29 then C1 to C29
And then ship out product
Inventory! Finished goods and WIP!
 Where is the waste
 If cycle time of A, B, C is 10. 20 & 40 minutes

 Big gaps in required operators & machines!

Minimize manpower & inventory
Sequencing
Simulate
 Exercise 1: # of cycle
quantity/cycle
1. Determine the number of cycle per day for this set of products that achieves fairly
level production?

Daily Daily Unit per Nearest multiple of Units

Product
Quantity cycle smallest quantity Short
F 9
G 8
H 5 5/5 1 -
K 6
2. Determine the production quantity per cycle for this set of products that achieves
fairly level production? It is important to mention that the number of extra units
usually verify the number of units short.
Cycle 1 2 3 4 5

Pattern F(2)-G-H-K F(2)-G(2)-H-K F(2)-G(2)-H-K F(2)-G(2)-H-K F-G-H-K(2)

Extra Unit(s) F FG FG FG K
 Exercise 2: Batch size
1. Total 1-shift production time available (excluding break time) is 7.5 hours.
There are 2 shifts per day. Average downtime per day (not included change
over time) is 40 minutes. Determine the available time for changeovers per day
and the desirable number of changeovers per day

Part# Average demand Cycle time per Average change over

per day (pieces) piece (seconds) time (minutes)
A 250 40 50

B 350 45 50
Determine best batch sizes
 Exercise 3: Batch sequence

Model Monthly Output per shift Takt time

output (units) (units) (min)
A 9,600 240 = 9600/40 2’=480/240

B 4,800 120 4’

C 2,400 60 8’

Sum 16,800 420 1.14’

20 operating days in a month, 2 shifts (8-hour) operation per day

What batch sequence should be?
 Sequence schedule for
introducing models
 All product varieties can be produced according to the
average cycle time of all varieties as long as each
model’s cycle time is considered when determining the
sequence of each model
Lot production:
AAA…A (240 units) BBB…B (120 units) CCC..C (60
units) ?
Or ???
A (2’) B (4’) C (8’)  Smoothed production:
AABAABCAABAABC
 Sequencing method for the
mixed-model assembly line
1. Determination of a cycle time (synchronized as takt
time)
2. Computation of a minimum number of processes
3. Preparation of a diagram of integrated precedence
relationships among elemental jobs
4. Line balancing
5. Determination of the sequence schedule for
introducing various products to the line
6. Determination of the length of the operations range
for each process
Goals of controlling the assembly
line
(1) Leveling the load (total assembly line)
on each process within the line
 workload streaming line
(2) Keeping constant speed in consuming
each part on the line
 Sequencing model
(1) Work load streamingline
(2) Goal Chasing Algorithm for
sequence schedule
 Objective: maintain a constant parts consumption rate (create a
constant demand)
 Principle: minimize the difference between the ideal parts
consumption and actual parts consumption

Units
Ideal parts consumption

Actual parts consumption

Production sequence
Supporting smoothed production
 Flexible machinery
 Flexible manufacturing system (FMS)
 Why production leveling/
smoothing?
- Shorter lead time
- Smaller finished goods & WIP inventory
- Balance use of labor and machines
- Help determine personnel, equipment,
material needs
- Smoothed demand on upstream
process & suppliers
 HEIJUNKA
(Load smoothing)
 Objective:
Low WIP  small batch size  more number of
batches  more number of setups
(Heijunka requires scrambling of parts)
 Determine:

- Number of set up we can afford, considering total

time available per day, set up and processing time
- Batch size
 Objective : Maximize number of set
up/Change over
 Constraints:
Total available time for a process >=
Total setup time + total production time
 General procedure
• Find out daily demands, setup and processing times for machines for
each part to be scheduled, and calculate total available time per day
per process.
• Determine daily demand ratios.
• Decide on the production cycle sequence.
• Calculate total production time for each process.
• Calculate total available setup time for each process
• Calculate setup time for each process in terms of the variables, n`s.
• Construct the model with the objective function and the constraints.
• Solve the model and determine how many cycles are required per day.
• Schedule. Determine batch size, the number of production for each part
per day, determine how many parts are unaccounted for the daily
demand.
• Determine total time utilization at each process and schedule MTOs,
low volume production, and unaccounted parts, accordingly
 Review & H/W
40-Day Demand Daily Demand
Part A 1000 parts 25 parts
Daily demand
Part B 2000 parts 50 parts
Part C 3000 parts 75 parts
Setup Process

Machine Cut Mold Assembly Machine Cut Mold Assembly

Parts (min) (min) (min) Parts (min) (min) (min)
A 10 5 2 A 5 4 2
B 4 3 6 B 3 6 6
C 1 2 4 C 4 2 4

Time

Working time No. of shifts per No. of machine Total working time Total available time
Time per shift day allocated per day (min)
Efficiency
per day (min)

Cut 480 min (8 hrs) 2 1 960 95% 912

Mold 480 min (8 hrs) 2 1 960 100% 960
Assembly 480 min (8 hrs) 2 1 960 100% 960
 Part A Part B Part C
Daily demand 25 50 75
Ratio 1 : 2 : 3

AC B C B C AC B C B C AC B C B C etc…
Objective : Maximize nA+nB+nC
Constraints:
(SACut* nA) + (SBcut* nB) + (SCCut* nC )  Total
setup time available for Cut
(SAMold* nA) + (SBMold* nB) + (SCMold* nC )  Total
setup time available for Mold
(SAAssembly* nA) + (SBAssembly* nB) + (SCAssembly* nC ) 
Total setup time available for Assembly