SATYAMOPTESSA Scheduler for Semiconductor Fabs

1
SATYAM-OPTESSA Scheduler for Semiconductor Fabs

• Satyam- Semiconductor Subvertical Initiative

2
Semiconductor industry ..

• Immense pressure on the semiconductor
  manufacturers to
• - increase product variety
• -reduce delivery times
• This makes building to demand a complex
  proposition
• Fully automated manufacturing systems

3
Complex processes

• Substantial amount of investment involved
• (typical cost to set up a 300mm wafer
  fabrication unit-- around 2-3 billion ,around
  1-2 billion investment to set up a 200mm unit)
• Facilities typically consist of around 75 to100
  different tool types
• With 300 pieces of equipment
• High-volume, high-mix facility like a foundry can
  have hundreds of products getting built in a
  facility at a time

4

• Hence , need for an effective planning and
  scheduling system to manage the complexity.

5
Many important processes are dependant on the
schedule

• Customer Relationship Management
• Dealer Order Management
• Material planning
• Scheduling of feeder lines
• Vendor supply chain management
• Effective assembly line sequencing / scheduling
  system provides a competitive advantage to any
  company.

6
Optessa MLS Optessa MLP have
been designed for manufacturing companies
with following characteristics

• With product variety,
• Producing in volumes,
• On one or more plants,
• Each plant consisting of one more lines, and
• Where sequencing or ordering of jobs is important

7
The OPTESSA advantage
8
Scheduling Semiconductor Cluster Tools
with OPTESSA MLS SC

• Can be deployed and used on complex tools used in
  semiconductor manufacturing to generate an
  optimal schedule/sequence of operations

9
 

• Conceptual image for a Cluster Tool

10
Semiconductor tool scheduling with OPTESSA
MLS-SC

• System can be used to maximize throughput through
  the tool
• Scheduler will optimally co-ordinate the
  activities of transfer, processing of wafers on
  the tool
• Ensures that the right resources have to be
  available at the right time

11
Continued

• Responds in real-time to random occurrences such
  as tool failure, downtime, alarms etc.
• Generates an initial optimal schedule that takes
  into account the global state.
• Makes truly intelligent decisions that provide
  long-term benefits

12
Requirements/complexities considered in the
semiconductor manufacturing scheduling

• Rescheduling
• Re-entrant Flows
• Non Uniform Load
• Setup Times
• Downtimes
• Rework
• Rejection of Lots
• Hot Lots
• Matching of Tools

13
Continued.

• Auxiliary Resources
• Test/Monitor Wafers
• Send Ahead Wafers
• Trial Lots
• Processing Time-Windows
• Utilization of resources
• Integration with MES
• Manual Overrides
• Rules/Constraints for Scheduling

14
The Optessa Solution

• Optessa MLS Feature
• Optessa MLS enforces precedence, time
  window and priority relationships across job
  classes, and develops optimal schedules
  considering all aspects of re-entrant flows

15
The Optessa Solution

• Optessa MLS Feature
• The support of cycle times that vary by tool
  or process step is a standard Optessa MLS
  feature. The system also has native support for
  both batch processing and single lot/product
  processing tools. Batches can be split and
  reconstituted dynamically at every tool.

16
The Optessa Solution

• Optessa MLS Feature
• Setups/ Changeovers times that account for
  from / to process steps are supported by the
  system along with the requirement to minimize the
  number of changeovers

17
The Optessa Solution

• Optessa MLS Feature
• One of the many strengths of the system is
  its ability to perform re-scheduling quickly in
  response to changes in the scheduling
  environment, such as random tool failures. The
  speed of the underlying algorithms allows
  rescheduling to be done in real time

18
The Optessa Solution

• Optessa MLS Feature
• Optessa MLS system supports priorities if
  rescheduling is required. Secondly, rules can be
  defined to space lots to make allowances for
  rework

19
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system supports scrap rates,
  and appropriately adjusts required volumes and
  production counts to ensure that requirements are
  met

20
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system supports due dates as
  well as the notion of hot-lots system seeks
  to satisfy line on / line off requirements on the
  basis of user defined priorities.

21
The Optessa Solution

• Optessa MLS Feature
• Optessa MLS supports line preferences for
  each process step, as well as sequence dependent
  updates of these preferences using run time
  attributes

22
The Optessa Solution

• Optessa MLS Feature
• Optessa MLS has a built in inventory simulation
  constraint which can be used to model scheduling
  restrictions caused by limited passive resources,
  such as reticles and FOUPS

23
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system has extensive support
  for test wafers which can be either scheduled by
  the system, or locked in by the user, as
  appropriate. Optessa MLS also has a number of
  constraint types that can enforce minimum or
  maximum counts of test wafers in a given time
  bin.

24
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system currently supports
  the requirement of splitting up a lot, sending a
  sub-lot ahead for processing, then scheduling the
  remaining lot

25
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system currently supports
  the ability to change or define new process
  flow/route information for each job class /
  process step

26
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system can currently
  support this issue by using a notion of
  time-windows that are defined parametrically in
  the system. In particular, the spacing between
  process steps for a given lot can be controlled

27
The Optessa Solution

• Optessa MLS Feature
• Another strength of the Optessa MLS system is its
  ability to maintain high throughput through all
  or critical resources by appropriately
  distributing the loads through out the system
  the cost function can explicitly model
  utilizations, which can be maximized while
  meeting all requirements and production rules.

28
The Optessa Solution

• Optessa MLS Feature
• Optessa MLS is easy to integrate with MES and
  ERP systems. Optessa is an alliance partner of
  Satyam, a world-wide leader in MES integration

29
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system currently has
  extensive support of manual overrides, including
  the ability for a user to readjust or make
  appropriate changes to the generated schedule.

30
The Optessa Solution

• Optessa MLS Feature
• The Optessa MLS system already supports an
  extensive list of rules/constraints that can
  capture the full range of rules / constraints
  that need to be considered in scheduling.
  Further, the modular design of MLS makes it
  possible to add additional rule types with very
  little effort.

31
Some of the numerous benefits

• Reduce order to delivery lead time by more
  frequent and closer-to-delivery scheduling
• Eliminate in process buffers and control parts
  inventory
• Reduce costs by considering all constraints
  
• Reschedule quickly to respond to changes
• Eliminate the need for the significant time and
  effort previously expended by line supervisors/
  planners in refining a sub-optimal solution

32
Scenario planning reduces capital
expenditures inthe plant

• Changing situations (such as market demand or new
  model introduction, raise the question
• Can we meet this new demand with the
• existing infrastructure?
• Solver can test whether the existing production
  rules and new capacity limits can both be
  satisfied. Capital expenditure worth several
  million dollars can be avoided in some cases.

33
Re-sequencing improves response to
changing market demand

• Manufacturer is more responsive to the
  marketplace by taking orders more frequently
• Orders are scheduled almost immediately and the
  impacted suppliers notified quickly
• Any changes in situation (e.g., supplier delivery
  problems) are handled in real time, as local
  re-sequencing can now be done in a matter of
  seconds.
• Iterative planning cycle is dramatically
  shortened

34
Optimization results in reduced
dealer and finished goods inventories

• Shorter lead time directly translates into
  reduced finished product inventory in the
  pipeline
• and in dealer lots with consequent reduction in
  the need for incentives like discounts and cheap
  financing

35
Optimization of supplier
restrictions results in smoothed
schedule for suppliers

• Supply chain constraints such as the daily
  capacity of the suppliers are now considered by
  the solver.
• Since the orders placed with the supplier are at
  or near its capacity, the supplier does not need
  to build big buffers to handle the spikes in
  demand that might otherwise generate huge
  contractual penalties.
• By eliminating the need to cater to huge
  variances in order, there is a significant
  opportunity to reduce in-plant capital
  expenditures.

36

• THANK YOU