Comments on the quotes from the
fab300_whitepaper:
“Within
the software framework, the cell controller emerged as the translator that
filled in the gap between the central control system and the need to
control the hundreds of pieces of equipment.”
|
|
Event-based
process control Figure 2. Event-driven MES systems can increase the speed of learning cycles, as data from equipment is easily converted to information, which can automatically cause actions to be undertaken. |
“All of
these qualities will let us move beyond the operator-driven paradigm of the
last 20 years, to an event-driven paradigm that is emerging in MES II systems. Figure 2 depicts an event-driven methodology,
where the circle represents a complete learning cycle. The learning cycle can
involve any learning that takes place in the fab:
from optimizing a process result in a single piece of
equipment, to managing maintenance cycles. The key point is that data leads
immediately and automatically to information, with little or no need for
trained staff to bridge various software systems to create the information.
Likewise, the information will be organized to trigger an action that closes
the learning cycle.”
“The improvement in ROI is a direct result of increasing the rate of
process knowledge on the fab line. With faster
process knowledge, chipmakers will potentially be able to attain
production-level outputs much earlier than before, at correspondingly higher
yields. We are already
capable of embedding more data analysis in the equipment so they can judge
their own “self health” and issue alerts that describe maintenance
requirements. Instead of relying so much on test wafers to detect many
equipment drift problem, as we do today, we should be able to more fully utilize
all the data streams available in the equipment itself and correlate them to
process results. Such systems should also be able to manage “smart” maintenance
dispatching so that the maintenance/repair schedule reflects the actual current
health or state of the equipment (versus a fixed maintenance and repair
schedule drawn up once or twice a shift).”
Comments
from QWiKS:
We propose to
add a Knowledge Management System (KMS) section to the Event-based process control Diagram. Our patented KM system
will provide all the infrastructure and flexibility the next generation MSE II
needs in order to obtain and store process knowledge. By embedding the KM
system into the equipment, the equipment can transform the “self health awareness ” into “self intelligence”. Our patented GUI
rule-based interface allows the engineers to add their knowledge into the
equipment by filling in an Excel-like table form. Being web based, the rule
entry interface can be accessed from anywhere within the customer’s intranet
and easily be added to or changed and can be used as a learning center for new
engineers.
“One
of the benefits of integrating the defect review process into MES is to relate
deviation or failure to both real-time and event-related reference points, by
referencing failure to a specific machine, lot, training requirement,
origin/supplier of wafer, etc. This activity has huge potential to increase
OEE, because of its impact on tool availability and repair schedules. Figure 5 shows a typical defect review process
flow.”
Comments
from QWiKS:
In
addition to using a defect review process to reference failure to a specific machine,
lot, training requirement or origin/supplier of wafer, we propose that adding the FAKS WET
module will allow reference failure to a specific machine from the device physics point of
view as this can directly relate back to yield or product performance problems.
“Improving Fab Yield
One of
the most compelling economic promises of next-generation MES II technology is
its potential to significantly improve fab yield and
reach high yield much faster. Several areas will contribute to this goal,
beginning with moving from today's Statistical Process Control, which can
identify an excursion or problem within a period of time corresponding to its
sampling frequency, to Advanced Process Control (APC) which can potentially
carry out a multi-dimensional analysis to identify the root cause of a
deviation and make adjustments to the control limits themselves or the
equipment/recipe conditions.
Comments from QWiKS:
By working
with our current customers, we have added several rules to help the users
improve FAB yields. One of the examples is to develop rules, which will capture
single or double sites that fail on a wafer which have a potential of mis-processing associated with those wafers. The current
common criterion for our customer is testing 5 sites per wafer. If 3 sites
fail, then the wafers would be treated as a failed wafer. Using our Rule-based
KM system, it is very easy to add more rules in order to improve FAB yields.
Rules can also be added that directly relate to root causes. When another lot
is tested that has similar traits, the rule will automatically be checked and
the lot will be identified as failing by the root cause.
Getting to the Next Generation
This
vision of a next-generation system finally integrates the disparate factory
management/automation system components with intelligent equipment. It rests on
several key assumptions:
1.
Equipment that has greater intelligence – with embedded process control
elements that share data with a factory-wide system, and predictive maintenance
(self health) – and is GEM compliant. This intelligence must be designed to
contribute to the fab's business objectives.
2.Event-based work flow. No longer is the action of an operator the basis for
the control system’s structure and rationale. This is the meaning of
automation.
3.
Extensive access to data of high granularity. The system will use a system-wide
data model, shared by all components, allowing management to add or replace
components without disrupting the other components.”
Comments
from QWiKS:
We
believe that our KM technology can greatly enhance the equipment’s intelligence.
Our proposed FAB Knowledge-based Engineering Nerve system can work in harmony
with MSE II’s Event-based work flow. Our KM
architecture has the potential to incorporate data mining technology to further
improve the FAB yields and automate most of the day-to-day engineering
processes.
“SUMMARY
The
stakes involved in the 300 mm fab continue to
escalate in terms of capital costs and risk. The factory management/automation
system of the 200 mm fab is no longer adequate to
maximize our investment and minimize our risk. A new model is required,
reinventing both the role of the equipment and the capabilities of the factory
systems. These systems can no longer be totally divorced from the equipment and
vice versa. We are on the verge of implementing a total solution factory
control system that can give management all of the tools they need to increase fab efficiency and return on investment, called MES II. It
can contribute to a dramatic reduction in
ramp-up time for new fabs, through out-of-the-box
equipment integration into the MES, as well as fast tool performance
monitoring. It can improve OEE through integrated real-time scheduling,
maintenance and health monitoring, and tool knowledge. In addition, yield can
be ramped faster and possibly increased overall through integrated SPC/APC and
proactive tool control.
Increasing
Overall Equipment Effectiveness (OEE) in Fab
Manufacturing To find out more information about
increasing OEE in fab manufacturing by taking
advantage of Consilium’s MES II solution, please
contact:
Applied Materials, FPS Product Business Group
408.584.6189 phone
408.584.6130 fax”
Comments
from QWiKS:
We
believe that our FAKS architecture and design fit into the new model of next generation
MSE II system. By automating the FAB engineering data analysis process and
Wafer disposition, the missing link in today’s FAB MSE system is completely
fulfilled. We believe that this would be a great opportunity for both companies
to develop a MSE system which will really meet 300mm FAB’s
future needs and beyond.