In manufacturing, few things remain constant and true. Little’s Law
is one exception. The equation: Cycle time in days is equal to amount
of Work In Process (WIP) in units, divided by the daily output in units
is always true and meaningful. If, for example, you have a total of
1000 units throughout the work areas either being worked on, or sitting
around, and your output is 100 units per day, your unit cycle time is
10 days long. It is also true, although the equation itself does not
specifically point this out, that if you take action that results in a
buildup of WIP, and that buildup results in increased cycle time, you
have probably made a mistake in judgment.
Continuing on, it is
also true that if your WIP remains constant and your output is
decreased, your cycle time will go up. If manufacturing could maintain
close control over the cycle time of its product, from the time it
leaves the release point until it has completely finished, it could
safely predict to customers what they could expect in terms of
delivery. If the process were completely under control, and nothing bad
ever happened, there would be no problem in meeting requirements, if
those requirements were within manufacturing’s capabilities. Companies
that are able to control their cycle time and meet their requirements
are usually very successful manufacturers. They may not be successful
in other critical areas, such as sales and planning, however. Sales,
production control and executive management can make the best
manufacturing processes inept, simply because they often have the power
to do so, and just don’t operate from the proper production value set.
They make decisions and promises that are destructive to their
production processes in order to make their own measurements look
favorable. Then, when manufacturing can’t make the promises come true,
it is blamed for the failure.
Realistically, most companies
have unpredictable manufacturing processes. Problems occur every day
and some of those problems may have lasting negative effects. People
don’t show up for work, machines break, vendor problems and an
assortment of other problems, are all designed to lower the output
capability of the factory. When the output capability goes down, and
the WIP remains at constant or at increasing levels, the cycle time of
the product goes up. When cycle time goes up, and product doesn’t get
out of the door on time, the prognosis for success goes down
dramatically. When manufacturing runs into problems and the demand
still exceed its capabilities, it is sometimes possible to make up the
delta through overtime or by bringing in extra workers. When this
fails, and it often does, chaos reigns supreme! If a company is running
very close to its maximum capabilities, and troubles hit, pushing it
beyond its capability, the probability for disaster is high. When
troubles start, it is very likely that questionable business decisions
will be made that will make the problems worse.
The reason is
that management starts using values that are designed to make things
worse under the guise of customer service. Prioritize this order, push
that order through, and release extra WIP to make up the difference. If
they really knew what they were doing, they wouldn’t react that way.
Little’s Law gives us a set of manufacturing values that always work.
Built within the simple equation are rules that should never be broken.
These are values, and as such, they guide us, no matter how tough it
gets. In a business that relies on the happiness of its customers,
which is probably true for a majority of businesses, the reliability of
the product, and the timely delivery of that product, is essential, not
only to success, but basic survival. In manufacturing management, it is
important that you be consistent and predictable under a variety of
conditions. Always fall back to those basic values and you will
prevail. Take the short-term loss and fix the problems by addressing
your detractors. Meanwhile, stick to the values contained within
Little’s Law.
There are a number of ways that a company can
react to production problems. The most common way is to go directly to
the problem area and try to fix it as quickly as possible. That
approach is just fine, except for the fact that the problem is probably
limited to one or two departments, while the rest of manufacturing
conducts business as usual. Consider this scenario: Department C has
problems that have reduced its output rate by a third. Every other
department has normal output capability. Engineering and maintenance
are working feverishly to fix the problem, but the situation remains
the same for a week. In the meantime, product is released to
manufacturing at normal levels. Customers are unaware that problems
exist that would threaten their delivery dates. Department A and B
produce at their normal levels, but WIP is building up at Department C.
The WIP that has ac*****ulated has gone beyond the delivery due dates.
The customers are not notified of a problem until their product has met
or exceeded the due dates. It could have been predicted that the
product would bunch up at Department C. Little’s Law is clear on that
point. If the input is the same, and the output goes down, WIP will
most definitely build up at the bottleneck. It was known up front that
C could not keep up the pace until their problem was fixed. That
identifies them as the bottleneck. If that were known, why would you
want to send them more than they could handle?
One could defend the logic by saying that it was hoped that the
problem would be fixed quickly, so they took the gamble and continued
with business as usual, but that logic is flawed. The truth is that the
company took appropriate action as far as going after the problem is
concerned. After that, they made errors that will probably create other
problems both short and long term. They could have minimized the major
problem by following the principles of TAKT scheduling: react quickly
to problems or opportunities on the floor by adjusting the rate
downward or upward on a daily basis. Little’s Law tells you how much
you can raise or lower your output. Once product is released to the
floor, it is essential that the company do everything in its power to
keep it moving. If you know that it will not continue to move, and will
build up somewhere, it is better to slow or stop the release of
product. It is impossible to keep it moving if you continue to release
product in timed quantities that exceed the capabilities of the
bottleneck. Once the product queues up somewhere, bad things are
probably going to happen to it. The product is probably going to be
handled more often than it should, and increased handling very often
results in quality problems. In addition, it can also be damaged while
it is just sitting there waiting to be worked on. If product is damaged
while it is idle, it is seldom reported because the perpetrator either
doesn’t know it happened, or doesn’t want anyone to know that it
happened because of his or her mistake. The effects of heat or humidity
on dormant product can also cause quality problems on certain types of
product. Continuing with the scenario, even when the problem is fixed,
it will take Department C some time (total WIP divided by output rate)
to work off the excess WIP, and the additional problems that I spoke of
will crop up eventually, creating even more slow downs and further
degrading the output rate. What was originally a machine problem will
suddenly become Quality problems. Adding to that is the fact that
customers will start to voice their discontent because they want their
product. The more clout the customer has, or the louder it complains,
the higher the chance that someone that deals directly with the
customer will exert pressure on manufacturing to increase the priority
of certain product. Then there is the pressure that is placed on
manufacturing by senior management: Hurry up! Who’s to blame for this?
We’re in serious trouble! This type of pressure begins the downward
spiral. Suddenly, the department is overwhelmed with requests for the
prioritizing of certain “Hot” orders. The effect of this is a further
deterioration of the output capability of that department, because the
simple act of prioritizing has devastating effects on output
capability. The effect is equivalent to a downward spiral of output
capability to the point that manufacturing is soon incapable of
satisfying any of its clients. Meanwhile, back at the front, the other
departments are still pouring salt on the wound, meaning they are still
pumping new product into Department C at the original rate. They are
doing their thing. It’s not their fault that Department C is in
trouble. Basic corporate survival is at stake now. The best thing that
other departments can do is to show that they are not the bad guys.
“Cover Your Own Butt” becomes the battle cry. As a result, the problem
grows. The downward spiral has begun, and it will continue until this
company starts to work as a team, uses analysis instead of pressure,
and establishes across the board discipline based on sound production
values into its process.
Little’s law should have guided the
original response to the problem. By applying a few basic manufacturing
values, the problem would still have caused some pain, but the pain
needn’t have lasting and increasing effects. Never release product in
timed quantities that exceed the timed capability of any part of your
manufacturing line. Work hard to make decisions that take into
consideration the fact that you will lose capacity temporarily, but the
product that you are able to produce will have the same cycle time it
would have had without the problem. Work on the problem and accept your
temporary limitations, but do not try to continue releasing product
normally, as if you had no problem. Do not make decisions that increase
cycle time. How does that apply to our scenario? For the short term,
the proper response would have followed the same logic. The department
reports its problem promptly through TAKT. Maintenance and engineering
quickly and realistically determines the time line for the fix, and
informs manufacturing and Customer Support when the problem will likely
be fixed and production will be normal. Department C may or may not be
the normal gate of the factory. A one third loss of output in
Department C may equate to a much less factory output loss. This only
means that the adjustment may, or may not be one third. Whatever the
effect on the line is, it can be calculated through Little’s Law or
determined through TAKT Management.
Managers should know what
their capability is. Moving people around from one department to
another may further minimize the loss of capacity. The point is that
you are now in the mode of proper damage control. You are working to
increase output, but you are trying to maintain a low cycle time. As
soon as manufacturing determines the effect that the problem will have
on output (33% for 1 week), the next step should be to alter the input
rates to Department A to a level that is consistent with the known
degraded net output capability of Department C. If the problem is fixed
more quickly, it will easy to increase the output rate back to normal
levels within an hour of the fix. It is much less difficult than trying
to work off excess inventory for a week or more, or having to
re-release product because of quality issues. While the fix is being
worked on, the customers should be contacted to determine the proper
priority of product release. Level with them and find out which orders
are critical and which can be delayed. Customers usually react to this
kind of honesty positively. If every customer wants their product now,
which is unlikely if handled correctly, the answer is to treat them all
equally. Prioritizing will only make things worse unless you have an X
Factor of less than 5. While the problem exists, employees in other
departments can take time to do other things, like training and PM’s.
If there are extra people available, they could be used to make sure
that when the product is ready to ship that there is someone there to
process it. Once the fix happens, you use the overtime to catch up, not
work off excess inventory that has been sitting around. The company may
have lost ground on their monthly output, but they can probably make up
the difference. The big difference is that the orders that are produced
are shipped on time with no loss of quality, and they are now well
positioned for productive output. When the problem ends, Department C
is now well poised to quickly get back up to speed because there is no
large buildup of WIP in their area. The problems that accompany WIP
buildup aren’t there. The process is under control, and now the output
increases as it should, and the cycle time remains constant or is
improved. The original problem has still caused some pain, but the
damage was minimized. The science of manufacturing was used to solve
the problem, not the panic tactics of management.
Little’s Law was observed because the amount of WIP released was
consistent with the required cycle time and output capability. If a
company wants to increase their output, they don’t do it by increasing
the amount released to the floor, when those output levels can’t be
reached. They do it by finding ways of increasing the output of the
slowest department. They find the bottleneck and work to increase its
output. If they have done everything that they can and still can’t get
their output to equal required levels, they may have to buy more
equipment or hire more people. They don’t make promises that are right
at the edge of their capabilities because they know that something bad
will probably happen that may swing the momentum in the wrong
direction. If there is to be growth, it must be accompanied with proper
planning and execution. Little’s Law presents an equation on the
surface, but says much more. It says that product cycle time is of
prime importance. If you do anything to increase it you are probably
making a mistake. If you reduce cycle time, you should do so by
increasing your output capability, and adjusting the WIP to meet the
demand. You do not reduce cycle time by simply lowering the amount of
WIP on the floor.
Do
not operate on the edge of your capabilities or accept orders that are
at the edge of manufacturing’s capabilities. If you do, you are rolling
dice that will eventually come up losers. There are times when
releasing extra product to the floor would be advantageous, such as
creating a temporary Strategic Kanban for equipment scheduled to go
down for an extended period, but with the capability of catching up
quickly, once up. This would allow the department behind them to work
off the WIP in the Kanban while the equipment is being worked on, thus
giving the appearance of no down time at all. The cycle time would go
up, but the effect would be continuous output at required levels. Also,
temporarily increasing the level of one or more buckets of a Dynaban
because of a similar problem (You know what is coming, and you know how
long the situation will exist) could also save you some pain. Proper
management means preparing yourself for certain situations. Most
importantly, you are following proper manufacturing values and
observing the science of manufacturing. Many companies have strong
product lines, and have buyers standing in line, but fail as
manufacturers because they can’t meet the delivery dates. In racing to
fill the demand, they resort to manufacturing values that only bring
their capabilities down, instead of creating improvement. The orders
exceed the capability of the factory, but they take them anyway. They
are afraid that if they turn a customer away because they may lose them
forever. They forget about, or fail to even care about, the science of
manufacturing and start pressuring their employees to produce more than
they are capable of producing. Management is under the gun to make more
product and they pass that pressure on to the people that have to make
miracles happen to save the company. Manufacturing gets the brunt of
the pressure, and when you pressure your producers, you may get
negatives that you didn’t bargain upon. Production Control and
scheduling in most factories is operated as a stand-alone department.
They usually have IE standards that guide them in their order
acceptance criteria. Little’s Law means nothing to them, as it means
nothing to Sales. Each of these departments base their operations on
criteria created by other support departments, such as Engineering. If
it were true that every department was marching to the same drummer in
terms of understanding capabilities, there would be harmony between
Sales, Production Control and Manufacturing. It is seldom true unless
PC and Sales truly understand Manufacturing’s capabilities. For this to
happen, the standards have to be created through historical TAKT data.
TAKT scheduling is the only sure fire way of making sure that when a
promise is made, there is a very high probability that the order will
leave on time. It is the only way that the bottleneck issues will be
addressed because the bottleneck is right out there for all to
identify. You release according to the capabilities of the bottleneck,
and if that isn’t enough to meet demand, you go out there and increase
your capabilities.
When you have accomplished that and have
demonstrated that the increased capability is not a fluke, you start
releasing more. Little’s Law is never violated, and is used to
calculate needs. Excess WIP doesn’t create new problems and pressure is
put where it should be put - on fixing the problems. It is oriented
towards increasing capabilities. So, what does the LL equation say? It
says, quite simply that, in order to achieve an acceptable product
cycle time, a company MUST balance the amount of product released to
the floor with the capabilities of manufacturing. It says that if you
increase the amount released, you must have an increase in the amount
produced. If you don’t, it says that you will increase your product
cycle time. If you increase cycle time you have probably made an error
in judgment. If you decrease cycle time and increase output, you have
something to be proud of. It is all so simple, yet so easy to overlook.
Little’s Law has been around for a long time. For a few years, it was
deemed important just to reduce the amount of WIP on the floor. A lot
of people were promoted or given bonuses because they reduced inventory
by thousands or millions of dollars. All they did was reduce the amount
of WIP to a level that was consistent with their output, if they were
lucky. If they weren’t, they took too much work off the floor. There is
more than one part of the equation there are three, with Cycle time
being the prime number.