Kanban Card Calculations

1
Kanban Card Calculations
2
1. Production Kanban
3
1. Production Kanban
Parts consumed during one kanban cycle Average
demand x kanban cycle time Since the demand
is not constant, a safety coefficient (?) is
introduced to the above formula Parts consumed
during one kanban cycle Average demand x(1?) x
kanban cycle time As a rule-of-thumb, ? should
be kept less than 10 percent and should be
decreased as the demand becomes more leveled.
4
1. Production Kanban
The kanban cycle time is the time spent by a
kanban to complete a full cycle it involves six
distinct phases as numbered in the figure above.
They are 1. The production kanban is removed
from the full container when the container is
withdrawn from the final buffer for transfer to a
downstream process the kanban is then placed in
the kanban receiving post and waits there until
collection. 2. The kanban is transferred from
the kanban receiving post to the kanban ordering
post at the initial workstation. 3. The
kanban waits at the production ordering post
behind the other production kanbans (FIFO). 4.
The kanban is taken from the production ordering
post and attached to an empty container, the
process machine are set up, the quantity of parts
to process (equal to the container size) that was
indicated on the kanban are withdrawn from
theinitial buffer the parts are then processed
and placed in the container upon completion. 5.
The full container (with its kanban) is
transferred to the final buffer. 6. The
container waits until withdrawn by downstream
process (the cycle is then complete).
5
1. Production Kanban
Kanban cycle time kanban waiting time in
receiving post kanban transfer time
to ordering post kanban waiting time
in ordering post lot processing time
(internal setup
run time in process waiting time)
container transfer time to final buffer
container waiting time in final buffer.
6
1. Production Kanban
Example Average demand 100 parts/hour, Contain
er size 100 parts, Kanban waiting time in
receiving post 5 min., Kanban transfer time to
ordering post 5 min., Kanban waiting time in
ordering post 30 min., Maximum internal setup
time 5 min., Maximum processing time 0.25
min/part, In-process waiting time 10
min., Container transfer time to final buffer 5
min., Container waiting time in final buffer 10
min., Initial safety coefficient ? 0.
7
1. Production Kanban
Thus, Kanban cycle time 5 1 30 (5
0.25x100 10) 5 10 91 min.,
Rounded up to the nearest integer, we get 2
kanban cards. Based on 2 kanban cards the actual
kanban cycle time is
The difference between 91 and 120 minutes is
about 25 of the kanban cycle time thus there is
no need to consider an additional safety
coefficient.
8
2. Withdrawal Kanban with Constant Reorder
Quantity
9
2. Withdrawal Kanban with Constant Reorder
Quantity
The number of kanbans is similarly calculated by
the formula
The withdrawal kanban formulas are similar to the
production kanban formulas. Parts consumed during
one kanban cycle Average demand x(1?) x kanban
cycle time
The kanban cycle time is calculated as
follows Kanban cycle time kanban waiting
time in receiving post kanban conveyance
time to upstream buffer container
conveyance time to downstream buffer
container waiting time in downstream buffer.
10
2. Withdrawal Kanban with Constant Reorder
Quantity
Example Average demand 100 parts/hour, Containe
r size 100 parts, Kanban waiting time in
receiving post 5 min., Kanban conveyance time
to upstream buffer 10 min., Container
conveyance time to downstream buffer 15
min., Container waiting time in final buffer 10
min., Initial safety coefficient ?
0. Kanban cycle time 5 10 15 10 40
min. The initial number of kanban cards is
then set equal to one, and the actual kanban
cycle time is
11
3. Signal Kanbans for Lot production
12
3. Signal Kanbans for Lot production
Lot size and the position of the signal kanban
have to be determined to enable the regular
replenishment of the final parts inventory. The
lot size is determined as a function of the
average number of setups per day. For example, if
on the average the total setup time (external
internal) takes two hours, then the maximum
number of setups per day is on the average equal
to eight for a two-shift operation. The minimum
lot size can then be computed as follows
For small amount of demand variability we can add
the safety factor to the above formula to get
13
3. Signal Kanbans for Lot production
Example A machine that produces five different
parts. Therefore, it has on the average a maximum
of 8/5 1.6 possible setups per part. If the
demand for that part is 1000 per day, and the lot
size is 100, then the minimal lot size for that
part, when ? is equal to 10 percent is calculated
as follows Minimal lot size (1/1.6)x1000x110
688 units. The minimal lot size is the most
economic lot size (minimumWIP) however since
every container must contain 100 units we round
the number to the next hundred, 700. Thus, each
production lot will produce 7 container for each
setup. Therefore, this will be ordered each time
a lot production is initiated.
14
3. Signal Kanbans for Lot production
The next question is the position of the
production ordering signal kanban in the lot of
processed parts. In another words, as the
700-unit lot is being consumed downstream 100 at
a time, when do we give the signal for the next
lot of 700 to be initiated? This position is
determined by the desired level of safety stock
at the time a new lot is reordered. The safety
stock is equal to the average number of parts
that are consumed during the time interval from
when the lot is ordered to the delivery of the
final product to its stock location. The
following formula formalizes this. Production
signal kanban position Average demand x(1?) x
kanban cycle time This position must be expressed
in integer number of containers, thus
15
3. Signal Kanbans for Lot production
The kanban cycle time is defined as Kanban cycle
time kanban waiting time in receiving post
kanban transfer time to
ordering post
kanban waiting time in ordering post
lot processing time
(internal setup run time in process waiting
time) container transfer
time to final buffer
16
3. Signal Kanbans for Lot production
Example To demonstrate with the ongoing
example, the following quantities will be
used Average demand 1000 parts/day or 1.042
parts per minute, Lot size 700 parts, Kanban
waiting time in receiving post 5 min., Kanban
transfer time to ordering post 1 min., Kanban
waiting time in ordering post 30 min., Maximum
internal setup time 10 min., Maximum processing
time 12 sec./part, In-process waiting time 10
min., Lot transfer time to final buffer 5
min., Initial safety coefficient ?
10. Kanban cycle time 5 1 30 (10
(12/60)x700 10) 5 201 min. Production signal
kanban position 1.042x(1 10)x201/100
2.30 containers. This number will be rounded up
to 3.
17
3. Signal Kanbans for Lot production
The position of the material ordering signal
kanban is similarly determined by examining the
number of finished parts that are withdrawn from
inventory during the time delay from raw material
ordering to new lot production ordering. The
value of this time delay must be set so that the
raw materials are available at the machine when
the internal setup is to begin. The delay is then
equal to the material-ordering lead time minus
the production-kanban lead time from ordering to
machine setup. Material ordering lead time
material-kanban waiting time in receiving post
material-kanban transfer time to raw
material storage material-kanban
waiting time at raw material storage
time to withdraw material from storage
time to convey raw material to process Production
-kanban lead time production-kanban waiting
time in receiving post
production-kanban transfer time to ordering
post production-kanban waiting time
in ordering post Time delay Material
ordering lead time - Production-kanban lead time
18
3. Signal Kanbans for Lot production
Note that the value of time delay may equal to
zero or it may even be negative. If this is the
case then the raw material should be ordered at
the same time as the lot is ordered and the two
cards should have the same position. When this is
not the case, the difference in position of the
two signal kanbans is given by
19
3. Signal Kanbans for Lot production
Example Continuing with our example, the
following quantities will be used Average demand
1000 parts/day or 1.042 parts per
minute, Container size 100 parts, Material-kanba
n waiting time in receiving post 5
min., Material-kanban transfer time to raw
material storage 15 min., Material-kanban
waiting time at raw material storage 15
min., Time to withdraw material from storage 10
min., Time to convey material to machine 10
min., Production-kanban waiting time in receiving
post 5 min., Production-kanban transfer time to
ordering post 15 min., Production-kanban
waiting time in ordering post 15 min., Initial
safety coefficient ? 10. Time delay 5
10 15 10 10 - 5 - 1 - 30 14 min. Kanban
position difference 1.042x(1 10)x14/100
0.16 container. Since the difference must be
integer, we round the number to 1. Therefore,
the material ordering kanban is placed one
container above the production-ordering signal
kanban.
20
3. Signal Kanbans for Lot production
21
KANBAN GOLDEN RULES

• Do not move nonconforming parts to a downstream
  process
• Ensure that downstream processes withdraw parts
  from upstream
• processes in the correct quantity and at
  the right time
• Do not let upstream process produce more than
  the quantity of parts
• withdrawn by downstream process.
• Ensure that the production is leveled the
  underlying rule to insure leveled
• production is that the kanbans should
  always be processed on a first-come
• first-serve basis. Altering the card order
  will result in unnecessarily
• speeding up some parts and delaying other
  parts.
• Do not attempt to transmit large demand
  variations with kanban system.
• Balance cycle times for smooth production, and
  constantly improve cells
• and workstations.