38 ISE Magazine | www.iise.org/ISEmagazine
Even though they were separated by half a continent and a major
ocean, Alexandria Industries in Minnesota and Bosch Hinges in the
Netherlands were facing exactly the same problems.
Both companies produced a high variety of products in low vol-
umes – Alexandria made aluminum extrusions, and Bosch manu-
factured specialty metal hinges (see Figure 1) – and both had long
lead-times of six to eight weeks. Despite having this large window of time,
both organizations were struggling with late deliveries, which required daily
expediting efforts in attempts to get jobs out on time. For several years, the two
enterprises had tried a plethora of manufacturing and control strategies, but
none had alleviated this ongoing problem.
Then, although separated by some 5,000 miles, management at each com-
pany independently decided to implement POLCA, an acronym for paired-
cell overlapping loops of cards with authorization. At that time, POLCA was a
relatively novel idea. Within a few months, both companies had reduced their
quoted lead-times by more than 50 percent, and even with this smaller time
window they were able to achieve near-perfect on-time delivery.
With the shorter lead-times and excellent delivery record, combined with
their reduction of expediting activities and costs, Alexandria and Bosch expe-
rienced significant increases in sales as well as profitability. And thus it was that
these two companies joined the ranks of the pioneers of POLCA in both the
new world and the old one.
POLCA picks the right job at the right time
So why was POLCA able to achieve in a relatively short time what these two
companies had been unable to accomplish for many years? The answer to this
question contains important insights for the industrial and systems engineering
community, and we will build up to the complete answer by the end of this
article.
One particular insight is worth mentioning right off the bat. ISE profession-
als have expertise in the design of systems, as clearly indicated in the profes-
sion’s name. These systems are often based on sophisticated mathematical mod-
els developed by the industrial and systems engineering community. However,
implementation experiences with POLCA have shown that another area of
ISE’s expertise is critically important – namely the human factors and socio-
technical aspects of a system’s design. This article will show that POLCA is
actually a fairly simple system, and the success of POLCA is strongly linked to
the “people side” of system design and implementation.
The terms used in the full name of the tool (again, paired-cell overlapping
loops of cards with authorization) are based on specic features of the system,
explained soon.
E
POLCA helps jobs dance
through your factory
Visual control system dramatically cuts lead-time by matching
cards against production capacity
By Rajan Suri
October 2018 | ISE Magazine 39
40 ISE Magazine | www.iise.org/ISEmagazine
POLCA helps jobs dance through your factory
First, lets understand what POLCA
does for a company. POLCA is a card-
based visual control system that manages
the flow of jobs through the shop floor:
At each of the main operations, it con-
trols which job should be worked on next
in order to meet delivery targets. Speci-
cally, it ensures that upstream operations
use their capacity effectively by working
on jobs that are needed downstream,
while at the same time preventing ex-
cessive work-in-process (WIP) buildups
when bottlenecks appear unexpectedly.
POLCA is particularly effective in
high variability environments where the
variability arises for external reasons such
as unpredictable demand, many different
types of orders and dynamic changes in
customer requirements, as well as inter-
nal reasons such as changes in equipment
and labor status and other day-to-day
conditions on the factory floor.
The context for POLCA stems from
the fact that in recent years manufactur-
ing companies have experienced signi-
cant market pressure to reduce lead-times
and be more responsive to customers.
Along with this, order sizes are getting
smaller, and companies are seeing more
orders for individually customized prod-
ucts.
To reduce lead-times significantly,
companies are reorganizing their shop
floors into cells. A cell is a collection of
equipment along with a team of people
responsible for the cells operation. So,
to satisfy the demand for numerous dif-
ferent small-batch orders or customized
products, companies often need to route
jobs through various combinations of
cells. This leads to the first feature of POLCA, which we will
explain via a simple manufacturing example.
Loops of cards between
pairs of cells
For our example company, let’s say certain types of jobs need
to go from the fabrication cell “N” to an assembly cell “S” and
then to a testing cell “T.
In POLCA, if jobs flow between any two cells, these cells
are connected by a POLCA loop. Thus, as shown in Figure 2,
there will be a loop from N to S and one from S to T. Each
loop contains a number of cards called POLCA cards that cir-
culate in the loop. The cards are specific to each loop and are
labeled based on the origin and destination cell.
For the two loops in the figure, they would be called N/S
cards and S/T cards. To support the visual nature of the sys-
tem, each cell is assigned a color, and the cards are color-coded
as also shown in the figure. This allows people to quickly as-
sociate a card with the origin and destination cells.
The function of the POLCA cards is to signal the availabil-
ity of capacity at destination cells. This is how the cards work.
When cell N is scheduled to start a job that next is destined for
cell S, then the cell N team needs to have an N/S card avail-
able in order to launch the job into their cell. After the job is
FIGURE 1
Unhinged deliveries
Bosch Hinges was having a hard time getting their specialty metal hinges, some of
which are shown here, out the door in a timely fashion.
FIGURE 2
Cards against capacity
POLCA cards travel in loops between pairs of cells. The color-coded cards help regulate
workflow and manage capacity.
Fabcell
N
Assemblycell
S
Test cell
T
S/Tloop
N/Sloop
October 2018 | ISE Magazine 41
launched, the N/S card is kept with this
job or with its paperwork to signify that
the card is associated with that job.
When cell N completes the job, it sends
the job along with the N/S card to cell S.
The decision-making at cell S regarding
when it can start this job will be discussed
below. For now, let’s say cell S does start
this job. It also keeps the N/S card with
the job. Only when cell S finishes work-
ing on this job does it send the N/S card
back to cell N.
This explains the first key feature of
POLCA – the card coming back from
cell S essentially conveys the message to
cell N: “We finished one of the jobs you
sent us; you can send us another.” In oth-
er words, a returning POLCA card signi-
fies the availability of capacity in a given
downstream cell.
If you are already familiar with kan-
ban – another card-based control system
with its roots in the Toyota Production
System and now a key part of lean manufacturing – you might
have noticed an important difference between kanban and
POLCA.
Kanban is an inventory signal. When a quantity of parts is
used up, it tells the previous operation to restock that inven-
tory. However, as just explained, POLCA is a capacity signal.
This difference is a major reason why POLCA works better
than kanban for low-volume and custom production. Speci-
cally, kanban requires you to carry inventory at many interme-
diate stocking points, but for low-volume parts this becomes
inefficient and wasteful. And for custom-engineered parts you
can’t carry inventory in the first place – you don’t know what
exactly to manufacture until you have received the order and
the custom engineering has been completed.
But neither of these situations is a problem for POLCA. For
low-volume parts, you will see from the next section that you
only make parts as needed based on demand. And for custom-
ized parts, since POLCA allows routing between any sets of
cells, the routing for a completely new part is treated no differ-
ently than that for a part that has been made before.
Authorization list and decision rules
So how does a cell decide which job to start next?
Based on the due date of a job, manufacturers calculate
authorization dates for each of the cells in the job’s routing
by backward scheduling from the due date using the planned
lead-times for each cell. If the company is using an MRP (ma-
terials requirements planning) system, this is a standard feature
of MRP. Otherwise this can be done easily with a spreadsheet.
These dates are similar to “start dates” for people who use
MRP, but there are some additional rules to follow, which is
why we call them authorization dates instead.
Through sorting the data for all the jobs, manufacturers ob-
tain a list of jobs yet to be processed in each cell. This is called
an authorization list. While this also resembles the “dispatch
list” in an MRP system, there are important differences as you
will see below. The authorization list only contains jobs that
have not yet been launched into this cell, sorted by authoriza-
tion date.
The first rule in POLCA is that jobs with a date of today or
earlier are authorized, and jobs with a date of tomorrow or lat-
er are not authorized. Only authorized jobs can be launched.
The list also displays the “next cell” where each job will go
next. This leads us to the main decision-making logic used in
POLCA.
Despite the long acronym, the operation of POLCA is actu-
ally quite simple. When a team is ready to start another job in
its cell, the team just needs to answer three simple questions:
(1) What’s the next job on the authorization list? (2) Has the
material for this job arrived from the upstream cell? (3) Do we
have the right POLCA card for this job?
For example, if the team is in cell N and the next cell for
this job is cell S, then the team needs to have an N/S POLCA
card available. If there is an authorized job on the list and the
answers to questions two and three are “yes,” then this job is
launched into the cell, and the team can start working on it.
If the answer is “no” to questions two or three, then the team
goes back to question one and looks for the next job on the list.
FIGURE 3
Keeping it visual
This POLCA board is used in cell U at Bosch Hinges.
42 ISE Magazine | www.iise.org/ISEmagazine
POLCA helps jobs dance through your factory
In the event that all the authorized
jobs receive a “no” in questions two or
three and there are no more authorized
jobs, the team will not launch any other
production jobs into the cell (even if the
material has arrived and these jobs are in
the schedule for the near future). Instead,
it will engage in activities such as con-
tinuous improvement projects. We will
discuss this later in the article.
To help systematize the answer to
question three, each team has a bulletin
board where it organizes the POLCA
cards available for its cell. This is called
a POLCA board. Figure 3 shows an ex-
ample of a POLCA board used at Bosch
Hinges.
Later, we will explain the benefits that
derive from the combination of using the
POLCA cards, authorization list and the
above decision rules. For now, let us complete the description
of POLCA by explaining the last key feature of the system.
Overlapping loops at intermediate cells
Let’s return to the job that went from cell N to cell S, and lets
say it needs to go to cell T next.
Each cell must use the POLCA decision rules (the three
questions), so when the job arrives at cell S, the team must
rst wait till this job is the next authorized job (question one
above). Since the job has arrived at cell S, the answer to ques-
tion two is “yes.” For question three, the team needs to have
an S/T card available. If it does, the S/T card is allotted to this
job and it is launched into the cell.
Remember that the job arrived with an N/S card, and that
card will not be sent back to cell N until the job is finished.
So while it is being worked on in cell S, this job will have two
POLCA cards with it.
Figure 4 shows an example of a job in process at Bosch
Hinges along with its two POLCA cards. Except in the first
and last cells in a routing, a job will always carry two POLCA
cards when it is being actively worked on in any cell.
Continuing with our example, when the job is completed
in cell S, the N/S card will return to cell N, and the job along
with the S/T card will go to cell T. And then the same logic
will repeat at cell T and so on through its entire routing. Note
that at the last cell in the routing, there is no “next cell,” so
question three does not need to be answered.
Operational benefits of POLCA
POLCA has been implemented in many different types of
companies and industries. After implementing POLCA, these
companies have experienced numerous benefits. Now that
you have seen how POLCA works, we can describe these ben-
efits in more detail; knowing how the system operates enables
you to understand why the benefits occur.
First, no complex software implementation is required.
Shop floor control software such as finite capacity scheduling
and manufacturing execution systems can be expensive and
take time to implement and customize for a given company.
Factories have been able to implement POLCA quickly and
without a lot of expense.
Second, POLCA ensures effective use of capacity while
also reducing congestion. Let’s return to the job going from
cell N to cell S. Lets say it is the next authorized job, but the
cell N team does not have an N/S card. Suppose there are
five cards in the N/S loop, as shown in Figure 2. (POLCA
theory provides a formula for calculating the number of cards
in a given loop.) The fact that the team does not have any
N/S cards means that there are five jobs already at cell S,
or on their way to cell S, and none of those jobs have been
completed by cell S.
Working on yet another job would result in more jobs pil-
ing up for cell S. Instead, the POLCA rules would make the
team skip this job, and let’s say the next authorized job does
have the appropriate POLCA card available, then the POLCA
rules would redirect cell N to work on this job. Hence the
team would work on a job destined for a cell that could benefit
by receiving this job.
Thus, we see that the rules accomplish two important goals:
Avoiding more congestion where there’s a bottleneck and en-
suring that upstream cells work on jobs that will continue to
flow downstream instead of working on jobs that will just end
up sitting at a bottleneck.
Next, POLCA works well for high-mix, low-volume and
FIGURE 4
Cards can carry your production
This shop packet at Bosch Hinges shows two POLCA cards associated with a job
while it is in cell K. The job came from cell U carrying the U/K card on the left, and
since it is going to cell V next, it needed the K/V card on the right in order to be
launched into cell K.
October 2018 | ISE Magazine 43
customized production. Since demand is moving more and
more in this direction, it is important to have a system that can
deal with this environment.
Materials requirements planning (MRP) systems have dif-
ficulty coping with the variability in a high-mix, low-volume
company. As documented in the literature, the MRP logic can
result in lengthening lead-times, late deliveries and an increas-
ing cycle of rush jobs.
Also, kanban systems, while simple, are not appropriate for
low-volume or custom production as explained above. In-
dustry case studies show that POLCA has provided excellent
results in such environments by eliminating expediting and
achieving near-perfect levels of on-time delivery.
POLCA is a visual system. There is strong belief in manage-
ment circles today that visual systems can be very effective in
keeping day-to-day operations on track. Simple visual systems
are also in keeping with sociotechnical systems design. The
POLCA system aligns well with these principles.
The system builds on existing structures and systems. POL-
CA builds on the cell structure that is becoming more widely
used, plus it supports better coordination between cell teams.
While POLCA does not require an MRP system, if compa-
nies already have an MRP system in place, POLCA seamlessly
complements this system.
POLCA rules provide a formal framework for continuous
improvement activities. As explained in the rules above, if all
the authorized jobs receive a “no” in step two or step three,
the team will not launch any other production jobs. Instead,
it will engage in improvement activities such as putting time
into cross-training, studying a setup in order to come up with
ideas for setup reduction, engage in preventive maintenance
tasks and so on.
Today’s manufacturing strategies place great importance on
continuous improvement activities, and such activities form
the core of jobs for many ISE professionals. But the truth is, if
we just keep launching jobs to keep people busy, when can a
team have time to engage in any of these improvement activi-
ties? When implementing POLCA, teams can be coached that
whenever their production activity is stopped due to POLCAs
decision rules, they can use these periods of time productively
for improvement.
And finally, POLCA builds on the capabilities of humans
in the system. Many so-called “optimum” scheduling systems
lose sight of the fact that such systems are deployed by people,
and these people need to buy in to the system and support it.
This oversight has resulted in ineffective or even failed imple-
mentations.
In contrast, experience with many companies has shown
that people on the front line like using POLCA; they find it
simple and easy to understand, support its implementation and
help to ensure that it continues to work well.
Success through simplicity
Sophisticated systems can fail in their implementation if they
don’t get buy-in from people on the front line. Industry appli-
cations have shown that POLCA accomplishes with simplicity
what complex systems offer but are often unable to achieve.
Having the full support of shop floor personnel is key to
POLCAs success. Feedback from one company included the
statement, “It’s the folks on the shop floor that are singing the
praises of POLCA.
The final proof of success, of course, comes from the
bottom-line results for companies that have implemented
POLCA: Shorter lead-times, significant savings in expedit-
ing activities and costs and an excellent delivery record. As
stated by Fried Kaanen, president and CEO of Bosch Hinges:
“Implementing POLCA was revolutionary for our company.
Now our delivery time is revolutionary for our customers.
Rajan Suri is emeritus professor and founding director of the Center
for Quick Response Manufacturing (QRM) at the University of
Wisconsin-Madison. He is author of several books on QRM and re-
cently published the book The Practitioners Guide to POLCA:
The Production Control System for High-Mix, Low-Vol-
ume and Custom Products, which includes numerous POLCA
case studies.
Shorter lead-times satisfy customers
Customers, whether they are end-users, upstream departments or other businesses,
are demanding shorter lead-times across the board. As described in the main story,
POLCA can help.
Writing for IndustryWeek, Larry Fast also suggests that enterprises check that their
major suppliers have plans to reduce cycle times in their plants – the quickest way to
reduce lead-times. If they don’t, perhaps you can volunteer one of your resources to
work with a supplier engineer to find a suitable corrective action.
A 35-year veteran of the wire-and-cable industry, Fast also suggested that
organizations check with their suppliers’ flex capacity options, which would indicate
how quickly they can respond to increased demand.