|
Other Abstracts
|
 |
|
FabTime Cycle Time Management
Newsletter Abstracts - Volume 7 (10 Issues)
In our main article this month, we have a special
year-end write-up of industry resources on cycle time
management for wafer fabs. We were inspired to put this
together by various inquiries from subscribers about
where to get started in learning about fab cycle time,
and what to do to start a cycle time improvement project.
People asked us about papers, conferences, software
products, links, tutorials, and the like. Naturally,
many of the resources that we have to suggest come from
FabTime’s own website, since we have been working in
this area for seven years. But, in our travels, we’ve
also come across other useful industry references,
books, conferences, and sources for papers, and we
wanted to share some of those with you. We hope that
you find this article useful. We welcome your feedback
and additional suggestions.
In this issue, we have a FabTime software tip of the
month concerning use of a new priority-based alert. We
also have two subscriber discussion questions – one about
integrated metrology, and one about staffing models.
We’re pleased to introduce both topics, because
subscriber discussion has been rather sparse lately, and
we hope that some of you will choose to contribute. We
also have a response to last month’s article about “7
Things You Should Know about Wafer Fab Cycle Time.”
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month was discuss seven things
that we believe people need to understand in order to improve
wafer fab cycle times. We focus on high-level items, areas
through which fab management can have a significant influence.
These include utilization, tool redundancy, equipment downtime,
product mix, staffing levels, access to performance data, and
selection of metrics. There are many other factors that
influence cycle time in wafer fabs, too, of course (batching,
setups, hot lots, and holds, just to name a few). But we think
that the areas highlighted in this article are worthy of your
time and attention. We welcome your feedback.
In this issue we also have a FabTime-related announcement,
complete with a glowing endorsement from one of our newer customers.
We also have one last conference announcement for the year.
Our software user tip of the month concerns hiding chart
legends on your FabTime home page. We have one subscriber
response to last month’s article about lot release policies.
We’re hoping that subscriber discussion levels will ramp back
up in the future.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month we discuss several ways
that operating practices in fabs contribute to arrival
variability, and hence to cycle times. These include
releasing lots into the fab in large batches, forcing
full batches on lightly utilized batch tools, and using
carts for manual transportation of lots between steps.
Each of these practices directly increases fab cycle time.
We recommend relatively low-cost changes that, if
implemented, can help to smooth the flow of WIP through
the fab, and make cycle times lower and more predictable.
In this issue we have two announcements related to
industry conferences. Our FabTime software user tip of
the month is about how to look at shipped lot cycle times
for individual lots in more detail. We have no subscriber
discussion in this issue, but we do welcome your questions
or comments for future newsletters.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month we provide some ammunition
to the many people who have asked us for help in estimating
financial benefits from cycle time improvement efforts. The
dollar benefits from cycle time improvement are not as
easy to estimate as those from utilization improvement.
However, there are several clear benefits that stem from
variability reduction and cycle time improvement. If we
can reduce variability in the fab, we have the option of
squeezing the existing capacity buffer, and getting some
extra throughput out of the same toolset. Alternatively,
if we reduce variability in the fab, we can reduce cycle
time. Cycle time reduction is tied to several other
benefits: improved line yield, decreased WIP carrying cost,
decreased cost of engineering change notices, decreased
risk of obsolete inventory, and increased revenue from
time to market pricing premiums. In this article, we
review the first two of these benefits in detail, with
numerical examples, and include highlights of the other
three.
In this month’s issue we have a response from
newsletter subscriber MS Ham to last month’s question about
making dispatching decisions that account for downstream
batch efficiency. Our development team has also been
busy, and our FabTime software user tip of the month
concerns a new method for configuring flexible goals.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our relatively short main article this month,
we discuss a fundamental conflict in wafer fabs: the
pressure to simultaneously increase tool utilization,
while decreasing cycle time. As regular readers of this
newsletter know, utilization is one of the main drivers
of cycle time at the tool level. As utilization increases,
cycle time increases non-linearly, becoming very large
for tools with the highest utilization values. Despite
this fact, fabs are under cost pressure to increase
utilization, so that they can get more throughput out
of the same toolset. In this article, we discuss two
ways to resolve this conflict.
In this month’s issue we have an announcement
about my being named to the Supplier Advisory Board
for the Fab Owners Association. The FOA is a corporation
of semiconductor fab owners and associates who meet to
discuss common manufacturing issues, and to combine
strengths and resources. I recommend that you look into
it, if you are an independent device manufacturer.
We also have announcements for two upcoming industry
conferences.
This month’s FabTime software user tip of the
month describes how to use our new Forecast Outs Lot
List chart. In our subscriber discussion forum we have
a response from a subscriber about an attempt to implement Kanban cards in a wafer
fab (in response to last month’s article about lean
manufacturing). We also have a new subscriber question about making active dispatching
decisions to improve downstream batch efficiency. We
welcome your feedback regarding these topics, or your
other manufacturing-related questions.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month, we revisit a topic
that we last touched in the very first year of publication
of the newsletter: lean manufacturing and wafer fabs.
Lean manufacturing is a philosophy focused on reducing
waste and developing a more flexible and efficient
manufacturing process. This philosophy has application
to many types of manufacturing environments, including
wafer fabs. However, many of the lean manufacturing
implementation techniques currently in use were developed
for automotive manufacturing, and are problematic when
applied to wafer fabs. This does not mean that the
underlying principles of lean manufacturing can’t be
applied to wafer fabs - merely that we have to look at
the underlying techniques, and focus on more relevant
industry-specific implementation methods.
In this month’s issue, we have an announcement
regarding the IEEE Robotics & Automation Society’s
Technical Committee on Semiconductor Manufacturing
Automation. They have a useful new website available.
Our software user tip of the month is about setting
alerts for individual lots at specific operations,
highlighting two new alerts requested by a FabTime
customer. In our subscriber discussion forum, we have
a response to last month’s main article about cycle
time variability. We’re also pleased to welcome 11 new
companies and universities to the subscriber list.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
This month we have an announcement about new
members of the Fab Owners Association, of which
FabTime is an associate member. Our FabTime software
user tip of the month is about generating a lot
comments report (to display all of the MES comments
for a particular lot). We have subscriber discussion
related to last month's question about process cycle
efficiency, and about cycle time variability.
The subscriber discussion questions about cycle
time variability, in conjunction with a discussion
that we’ve been having with one of our
customers, inspired us to write this month’s
main article about cycle time variability (or the
distribution of cycle times). We first briefly discuss
benefits of and methods for tightening the distribution
of shipped lot cycle times. We then review a couple of
metrics for tracking variability within the fab, with
emphasis on understanding the impact of this variability
on overall cycle time distribution. We next describe
several methods for tracking and reporting shipped lot
cycle time distribution, including a new metric similar
to the A20/A80 availability metric, which we have called
CT20/CT80. We believe that this metric can help fabs to
better understand, and hence to improve, the distribution
of lot cycle times.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month, we summarize our
current recommended set of metrics for managing and
improving cycle time in a wafer fab. Metrics are an
indication of management priorities. Fabs that are cycle
time focused need metrics that drive cycle time
improvement. Such metrics should highlight current
opportunities for improvement, and, if followed,
should lead to improvements in shipped lot cycle
time. In this article, we discuss several metrics
that we believe are helpful for cycle time
improvement. We have categorized these metrics as
open lot, closed lot, shift-level, and tool-level,
and have briefly defined each one.
This month we have an announcement about the
newsletter receiving its 2000th subscriber. We also
have an announcement regarding future newsletter topics.
Our FabTime user tip of the month describes the use of
the new “stripe” control to apply a target band across
a chart. We also have subscriber discussion regarding
the continuing topic of M-Ratio (maintenance ratio),
by James Ignizio, and about a cross-fab metric called
Process Cycle Efficiency.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month we return to the ever-popular
topic of operators. Specifically, we discuss some of the
ways that operators can introduce variability into the
fab and suggest metrics to capture and reduce this effect.
It’s a relatively short article, in honor of our shortest
month (and because of the extensive subscriber discussion
in this issue).
This month we have two FabTime announcements, one
about our new associate membership in the Fab Owners
Association, and another concerning our enhanced lot
dispatching module. We also have a call for papers for
a special session on semiconductor manufacturing at the
upcoming IEEE Conference on Automation Science and
Engineering. Our FabTime software user tip of the month
is about displaying zero-value objects on the move and
WIP pareto charts. This month we have quite a bit of
subscriber discussion, with multiple responses to last
month's article about running development lots in a
production fab. We also have multiple responses to
James Ignizio’s description of M-Ratio from last month’s
subscriber discussion forum.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
In our main article this month, we discuss the cycle
time impact of running development wafers in a production
fab. For many fabs, running development wafers in a
production fab is a necessary part of doing business.
However, the development wafers can have a negative
impact on production lot cycle times. In this article,
we discuss several reasons why development lots may
contribute to higher production lot cycle times. We
also trace each of these factors back to their effect
on the three fundamental drivers of fab cycle time:
utilization, variability, and number of tools per tool
group.
This month we have announcements about a new software
development cost estimation tool from WWK, and a
FabTime News and Updates website/blog that we are
beta testing. Our software user tip of the month describes
how to add a custom title to any FabTime chart. This
month’s subscriber discussion forum brings feedback
from James Ignizio regarding last month’s article on
operational recommendations for fab cycle time
improvement, as well as a new subscriber question
about understanding the relationship between staffing
levels and equipment utilization.
(
Return to newsletter subscription page to
subscribe and receive the current issue
free each month)
|