Industrie 4.0 or digitalization of manufacturing currently create uncertainty and unrest in the manufacturing industry as many players do not know when, how or whether a disruptive change in industry will happen. Many published high-level strategies stay vague and leave practitioners unsure what to expect. Breaking Industrie 4.0 down into tangible pieces and steps is necessary for transporting the vision into reality. In this paper we develop an assessment and roadmap for Industrie 4.0 in semiconductor manufacturing the FINCA model. The model covers semiconductor frontend and backend manufacturing. It was successfully applied and tested at one of Europe's largest semiconductor manufacturers, the Infineon Technologies AG. Results from the assessment are presented in this paper.
• Applied computing → Reference models; Enterprise information systems; • General and reference; • Computer systems organization → Embedded and cyber-physical systems;
Industrie 4.0, digitalization or digital transformation create a spirit
of optimism but also a high uncertainty in the manufacturing
industry. On a general level the three terms have the same meaning:
The introduction of digital technology into manufacturing. Many
consultancies and research institutions expect a high impact on
manufacturing by the so-called fourth industrial revolution.
Fraunhofer IPA estimates an average cost reduction potential of about
30% [
Industrie 4.0 is a set of contemporary automation and data
science technologies, as well as organizational paradigms for
manufacturing in the 21st century. The core of Industrie 4.0 are
CyberPhysical-Systems (CPS), which connect the physical and the virtual
world [
The high number of diferent technologies associated with Industrie 4.0 leads to the question of prioritization of diferent approaches at companies. In a fast moving field, with standardization still ongoing, companies are reluctant to make investments in new technologies. High-level strategies ofer little orientation as they do not get specific enough to derive concrete recommendations. The fear of investing into the wrong technology slows down innovation tremendously. Strategies need to be broken down into smaller parts to provide tangible steps towards the implementation of an Industrie 4.0 vision.
There are several assessments and roadmaps for Industrie 4.0 and digitalization available (section 3). Still, no framework can directly be applied to semiconductor manufacturing. Most assessments are general and not industry-specific which leaves room for interpretation and leads to subjective results of the assessment. Additionally, no framework is currently available which can be applied to semiconductor frontend and backend to compare the level of digitalization in these manufacturing steps.
In this paper, we present a framework for Industrie 4.0 in semiconductor manufacturing. The framework can be applied to frontend and backend production. It can be used as assessment and roadmap for further development of the manufacturing site. The purpose of the framework is • to foster a common understanding between Industrial Engineering, IT and Business on the existing capabilities, • to create a vision for further development in semiconductor manufacturing, • to identify gaps at manufacturing sites, • to support benchmarking between semiconductor manufacturing companies, and • to enable a fast assessment of acquired sites within integration projects.
In the section 2, we will describe the semiconductor manufacturing process and the diference between frontend and backend. In section 3, existing frameworks, assessments and roadmaps for Industrie 4.0 will be presented. We also highlight some existing roadmaps for the semiconductor industry. None of the existing frameworks ofers a detailed semiconductor specific assessment and roadmap which can be used for frontend and backend manufacturing. Therefore, we developed the FINCA Model which will be presented in section 4. In section 5, the model is applied to frontend and backend sites of the Infineon Technologies AG and results are discussed. In the conclusion (section 6), further research directions and applications are presented. 2
The semiconductor manufacturing process starts in the frontend.
Structures in the sub-µ m range are processed on raw wafers, which
are thin slices of crystalline silicon. The manufacturing process
requires a cleanroom as dust or other particles can destroy the
sub-µ m structures during the fabrication process. From a
manufacturing point of view, frontends are complex job shops (for a
detailled description see [
Semiconductor frontends are considered high-tech with complex processes and high levels of automation and digitalization. They are very capital intensive and mostly located in advanced economies.
After the frontend the wafers are brought into an intermediate storage facility, the so-called die bank. From the die bank the wafers are taken to the backend, the second and final manufacturing step. At the backend, the wafers are cut into separate dies. The dies are bonded to a leadframe, which connects the chip to electrical contacts on the outside of the package. After the bonding, the chips are packaged and sealed in order to make them robust against environmental impacts. The final product is now ready for sale.
In contrast to the frontend, the backend is traditionally a more mechanical and labor-intensive process rather located in low-cost countries. Latest backend technologies which comprises of assembly and final test became more sophisticated and more complex. 3
The Platform Industrie 4.0 released the Reference Architecture
Model Industrie 4.0 (RAMI 4.0) [
There are several Industrie 4.0 assessments and roadmaps
available [
For technology development and the continuous shrinking of
semiconductor devices (Moore’s law) the International Technology
Roadmap for Semiconductors (IRTS [
The increase in wafer size has always lead to substantial changes
in manufacturing engineering at semiconductor plants. However,
the switch to 450mm wafer-size has been delayed and is not
expected within the next 2-3 years [
Current initiatives mostly focus on the application of specific
technologies in semiconductor manufacturing without providing a
full picture. Here, the focus is on intelligent algorithms [
All in all, the existing frameworks lack scope, are too general in their recommendations or do not focus on digitalization. 4
The FINCA model is an Industrie 4.0 assessment and roadmap for the semiconductor industry for both frontend and backend manufacturing. It was developed at Infineon Technologies AG. The main properties are already encoded in the abbreviation FI-N-C-A: • Factory Integration (FI):
FI refers to all IT services necessary to run a semiconductor production. In some companies the responsible organization is called “Manufacturing IT” and can be under IT or a diferent central function, local factories or cluster management. Among diferent tasks, FI’s mission is to ensure standardization within the company. At Infineon Technologies AG, FI is under the corporate supply chain function and has the mission to standardize across regions and manufacturing levels while maintaining and even increasing capabilities of the manufacturing system landscape. • Normalized:
Capabilities are, wherever possible, independent from region, manufacturing levels (frontend, backend) and products. Ideally any frontend site can be compared to any backend site using the normalized capabilities. There are five levels for each category going from zero (no capability or no system to support paper/manual process) to four (capability implemented in professional IT system and used to the fullest extend in regards of industry standards). Each category can be split into several sub-categories that need to be assessed
Industrie 4.0 Level Level 0 Level 1
Level 2
Level 3
System decides what will happen
Anticipate what will happen
Understand what is happening
See what is happening Connect the data Automated decisions Prediction Knowledge Information Connection Pre-requisite Stabilization & Standardization individually and aggregated using a simple numerical average (no weighting). • Capability:
Focusing entirely on capabilities and availability (rollout %) of those capabilities in one location/sub location (whatever makes sense in terms of an existing homogeneous capability landscape). Out of scope are architecture, technology stack, service levels, application names, source of the applications (build vs. buy) and infrastructure. Applications are only used in an abstract way like application classes e.g. “Manufacturing Execution System (MES)”. Application roadmaps, stability and architecture changes are only considered if they add/remove capabilities. • Assessment:
The FINCA model has to be assessed and filled by the business owner of a site or sub-site, usually supported by business domain experts, FI domain experts and FI business analysts. Business process experts and FI business analysts are in charge to keep the normalization of all dimensions (the grid) up to date, so the comparison independent of manufacturing levels or region is always possible.
The FINCA Model consists of seven dimensions and several sub categories. Each dimension can achieve a value from level zero (low capability) to level four (maximum in terms of desired capability). An overview of the dimensions is given in Fig. 1. Every dimension is described more precisely with the number of sub categories that are to be rated during the assessment. While all levels are separately defined, they follow a general guideline with diferent maturity levels. The diferent levels are depicted as a knowledge pyramid in Fig. 2. The foundation of the pyramid is “Stabilization & Standardization” and goes up to “Automated decisions”: • Stabilization & Standardization: Process is according to standard and running stable. First, local data collection is in place. • Connection: Data sources are connected, standardized and can be accessed globally. • Information: From data to visualized information, e.g. KPIs and cockpits. System processes data to gain information and to create transparency. • Knowledge: Classification of events based on information that may lead to triggered actions or automatic generation of proposals for action. • Prediction: Predict future events by simulation, machine learning or complex mathematical/statistical models. • Automated decisions: Autonomous systems base their decisions on anticipated events and an awareness for their environment.
To score a sub category, certain criteria have to be taken into consideration. They are called diferentiators. Those diferentiators are specific features and their existence (or their extend) in a factory can be used to rate a capability.
For example, the diferentiator “tool connectivity” can be used to rate the APC/FDC (Advanced Process Control / Fault Control and Classification) capability of a site. The tool connectivity determines to a great deal the amount of data that is available in the first place to allow for process control and monitoring.
As some factories do not have a consistent level e.g. some lines have more automation capabilities than other lines in the same factory, the level of a sub category can be broken down into multiple rollout scenarios. As some machines in a factory have a better connectivity than others a coverage/distribution/rollout percentage factor has to be applied. For example, if 80% of a factory’s machine park has an availability of 50% of the critical parameters covered in APC/FDC (equals level four) and 20% is connected but has a coverage below 50% (equals level three), the overall rating for this sub category is (80 · 4 + 20 · 3)/100 = 3.8.
Not always all five levels are available, in that case only existing levels as per description have to be used.
Once each sub category has a calculated value based on the diferentiators and the distribution of coverage across the levels, the overall dimension level is to be calculated as the average (non weighted) of its sub categories levels.
In the following, the capability categories and their sub-categories are presented. An overview of the capability categories is given in Fig. 1. 4.1 • Deviation Management System • WIP Routing (Workflow, Lot Route, . . . ) • Exception Management (Workflow) • Subcon [External] / Inter Site [Internal] Management • Small Lot Size Mastering [Lot Size 1] • High Automation Load & Go • Experiment Management System for Sample and Engineering Lots
4.2
• Forecasting for Volume • Dispatching • Scheduling • Work Area Control • Capacity Planning
4.3
Process control automation has eleven sub-categories • Documentation & Documents • Dynamic Parameters • Check Sheets • Work-In-Progress Data • Sampling & Buyof • Recipe Handling • Process Time Window / N2 Cabinet • Statistical Process Control (SPC) • Statistical Bin Analysis/ Automatic Lot Release • Advanced Process Control/ Fault Detection and Classification • Metrology
4.4
Manufacturing data management has eight sub-categories • Master Data Systems Availability • Master Data Systems Change/ Release • Master Data Static Systems Accuracy • Master Data Dynamic Systems Accuracy • Operational Production Reporting • Aggregated Reporting • Data Analysis • Lot Release The definition of the levels is given in table 5, 6, 7 and 8. • Product (WIP)/ Device (Lot, Strip, Chip) Identification, Validation & Traceability • Production Material & Wafer Material Identification, Validation & Traceability • Tool Identification, Validation & Traceability • Carrier / Container Identification, Validation & Traceability • Equipment Identification & Validation • Operator Identification & Validation • Non-Productive Products / materials / tools [durables] /
Equipment • Unified Material Mapping • Split & Merge The definition of the levels is given in table 10 and 11. 4.7
Equipment automation has six sub-categories • Equipment Interface • Equipment Data • Automated Setup/Change Over • Equipment Health Monitoring • Maintenance • Input loading/ Output loading The definition of the levels is given in table 12 and 13. 5
The FINCA model has been tested by semiconductor production experts of the Infineon Technologies AG. The model has successfully been applied as internal benchmark. The results were used to identify best practices and lead factories in certain areas. Next steps for development of the sites could be identified.
As an example for the application of the model the aggregated results of one frontend and one backend site the Infineon Technologies AG are discussed. The aggregated outcomes are shown in Fig. 3. The axis have been rescaled, but still allow for a relative comparison and discussion.
The semiconductor frontend is relatively advanced in terms of
Industrie 4.0. Frontends of the Infineon Technologies AG have a
very high degree of automation. The Infineon site in Dresden is the
200mm-wafer-size frontend with the highest degree of automation
[
Workflow Automation
According to a McKinsey & Company analysis Industrie 4.0 ofers
just the right tools for these productivity gains in backend [
Digitalization in capital-intensive frontends has started in the early 1980s. The early introduction of Manufacturing Execution Systems (MES) has lead to legacy systems in production. The learning from the frontend MES could be applied to the backend where introduction started significantly later. This diference can be seen in the FINCA dimension Manufacturing Data Management: The frontend site scores relatively low, while this is a strong dimension for the backend. This reflects the efort at the backend sites in the recent years to introduce a solid foundation for digitalization.
The assessment has provided useful insights for the next steps at both sites. Best practices or tools at diferent sites could be identified and transferred to other manufacturing locations. 6
In this paper we presented an assessment and roadmap for Industrie 4.0 for both frontends and backends. The FINCA model has been successfully applied at Infineon Technologies AG. It has proven itself to be a useful tool at evaluation and roadmapping for future improvements.
With this publication the authors want to foster the exchange with science as well as other semiconductor companies. In science, the FINCA model can be used as guideline how semiconductor manufacturers envision manufacturing in the future. The FINCA model assists researchers to find open challenges and problems. New technologies and approaches from science can help semiconductor manufacturers to reach new levels of productivity and quality.
The authors want to use the FINCA model to exchange with other semiconductor companies on their vision of Industrie 4.0 for semiconductor manufacturing. Furthermore, the assessment can be used for manufacturing benchmarks with other semiconductor companies.
FINCA was developed for semiconductor manufacturing. Still, we think the general model is also valid for other manufacturing industries. It is especially suitable for job shop production systems with large amounts of standardized products. The authors invite other industries to apply and test the model in their scope and welcome the exchange of experiences with FINCA.
Thanks to all who contributed to the Framework: E Chui Geok, Frank Banzhof, Frank Lehmann, Lim Shaw Ming Daniel, Michael Foerster, Joerg Domaschke, Woi Teck Khiong, Walter Laure, HansJuergen Wimberger, Sim Wee Sien, Yeo Danny, Teh Min Kiap, Christian Knoell, Michael Brueggemann, Daniela Eknigk, Mathias Haeuser, Chan Wai Ling, Nina Trude-Kuschel, Andrew Low, Goh Kian Thong, Marco Tschemmer, Harald Heinrich, Tan Jee Liang Jeffrey, Karl Horst Hohenwarter, Klaus Sandtner, Dirk Loefelmacher, Gustl Kreuzberger, Ronald Bianchin, Torsten Quaas, Tong Soon Hock Adrian, Olaf Herzog.
A part of the work has been performed in the project Power Semiconductor and Electronics Manufacturing 4.0 (SemI40), under grant agreement No 692466. The project is co-funded by grants from Austria, Germany, Italy, France, Portugal and - Electronic Component Systems for European Leadership Joint Undertaking (ECSEL JU).
This work was supported as part of the joint undertaking “SemI40” by the German Federal Ministry of Education and Research under the grant 16ESE0074. Results and statements in this paper reflect the viewpoint of the authors. A
Level 0 Level 1 Level 2 Level 3 Deviation Management System Diferentiators: Detection [Auto / Manual], Containment [Auto / Manual], Release [Auto / Manual] Process to handle devi- Manual detection with Auto detection with de- Following harmonized ation is defined, paper auto hold, auto detec- fined action / contain- containment action recording of deviation, tion with manual hold ment for quality and through standardized no deviation system in (standalone), manual yield areas (auto hold) deviation flow at FE / place detection with manual BE hold Small Lot Size Mastering [Lot Size 1] Diferentiators: Data Transfer [paper, File Transfer], Visibility [Black Box, Sub Operation, Sub Step], Data availability FE: Full wafer cassette FE: No full wafer FE: Compound Lot, BE: FE: - , BE: Lead frame lot Die level lot size proprocessing, BE: Stan- cassette processing, BE: Sub Standard Lot size size process cess. dard lot size (e.g. 25 Standard lot size (e.g. (e.g. Magazine) Wafer) process for all 25 Wafer) process for processes certain processes High Automation Load & Go Diferentiators: Loading [Auto / Semi auto / Manual] Manual loading Manual loading linked Semi auto loading with MES linked with MES
Level 0 Level 1 Level 2 Level 3 Level 4 Experiment Management System for Sample and Engineering Lots Diferentiators: Number of capabilities (auto split/merge lot, recipe name and parameter overwriting, APC overwriting, . . .), Existence of an experiment management system Manual creation & re- Fixed Route upon Fixed route upon release, manual mainte- release, existing exper- lease some capabilities nance, fixed/static alter- iment management nate item (Route, Bill system, basic caof Material, Tool plan...) pabilities (routing, upon release, no experi- equipment/equipmentment management sys- setup/tools) tem Forecasting for Volume Diferentiators: Data Feed [Manual / Auto], Frequency, Scope [work center / line / factory], Method [Simulation / Mathematic Formula] FE: Simulation & mathematical optimization, BE: Mathematical optimization
Manual data feed, Manual data feed, daily Semi-automatic data Semi-automatic data weekly forecasting, forecasting, work center feed, 6hrs - 8hrs fore- feed, 6hrs - 8hrs forework center forecasting forecasting, manual re- casting, line forecasting, casting, whole factory, manual reporting porting manual Reporting auto reporting Dispatching Diferentiators: Integrated line control [one system apply to whole supply chain], Compliance [work center / line / factory], Flexibility [rules definition by Equipment / Work center/ line], Timeliness <50% Compliance >50% Compliance >80% Compliance >90% Compliance
100% Compliance [Fully Automated], real time, integrated line control, full flexibility Scheduling Diferentiators: Integrated line control [one system applied to whole supply chain], Compliance [work center / line / factory], Flexibility [rules definition by Equipment / Work center/ line], Timeliness, Data integrity, Scope [lot start / whole line]
Paper recording of System recording of System warning of cre- Automated predictive creation/update sched- creation/update sched- ation/update due, sys- creation/update ule (fixed time, volume ule (fixed time, volume tem stop of maintebased) based), system warning nance due (integrated to of maintenance due MES)
ation/update schedule based on capacity optimization (integrated to resource, tools, spare parts demand, WIP) Work Area Control [Radar] Diferentiators: Users [Operator / Supervisor / Engineers], Scope [work center, Line, Equipment], Information [4M Man, Machine, Method, Material], Timeliness, View consolidation [One View, Multiple, easy access, mobility] View of critical line con- Snap-shot dashboard Snap-shot dashboard(1 Real-time dashboard Real-time Dashboard (1 trol information at the (multiple views) of view) of critical line con- (multiple views) of view) of critical line conequipment critical line control in- trol information (all Sys- critical line control in- trol information (all sysformation (all systems) tems) formation (all systems) tems) Capacity Planning Manual
Single only manual
Work (Bottle
Center Multiple Work Center & Neck), Line, manual
manual and partial auto
Level 0 Level 1 Level 2 Level 3 Documents Diferentiators: Paper / paperless, Search for correct Doc, Revision management Paper documents on Paperless documents Paperless documents Paperless document on equipment, manual on equipment, manual on equipment, semi au- equipment, automated search for the correct search for the correct tomated search for the search for the correct document (standalone), document (standalone), correct document (non document (integrated manual control to manual control to integrated/standalone), one transaction), man- display the latest display the latest manual/automated ual control to display revision revision control to display the the latest revision latest revision Dynamic Parameters Diferentiators: Paper / paperless, Search for correct Document, Revision management Paper documents on paperless documents Paperless documents Paperless info on equip- Paperless info in system, equipment, manual on equipment, manual on equipment, semi ment, automated search automated search the search for the correct search for the correct automated search for for the correct info (in- correct info (integrated info (standalone), man- info (standalone), man- the correct info (non tegrated - one transac- - one transaction), autoual control to display ual control to display integrated/standalone), tion), manual control to mated control to display the latest revision the latest revision manual/automated display the latest revi- the latest revision control to display the sion latest revision Check Sheet (Reminder to check tasks, anti-mix, Setup Yield, Test Program) Diferentiators: Paper / Paperless / Online control, Validation paper check sheet with paper check sheet with paperless check sheet online control no validation validation, four eyes val- with validation, four idation eyes validation, defined ranges paperless check sheet with validation, four eyes validation, defined ranges, warning/hold if out of range
Diferentiators: Paper / paperless /online control, Validation paper WIP data collec- paperless WIP data col- paperless WIP data col- paperless WIP data col- online control [auto coltion lection lection with validation, lection with validation, lection of WIP data] defined ranges warning/hold if out of range Sampling & Buyof (Products) Diferentiators: Paper / paperless / nothing, Triggering, Sampling Type [Static / Dynamic] paper based, manual paperless, manual trig- paperless, automated paperless, automated paperless, automated triggering, static sam- gering, static sampling, triggering, static sam- triggering, static sam- triggering, dynamic pling, 100% sampling ifx sampling rate, execu- pling, fix sampling rate, pling, fix sampling rate, sampling, execution rate, execution [man- tion [manual] execution [manual] execution [automated] [automated] ual] Recipe Handling (Tester recipe, Handler recipe, Assembly Recipe) Diferentiators: Recipe Release, Recipe select / download, Recipe Validation [Body check] manual select from local semi auto select from lo- manual download of semi automated down- automated download m/c, manual adjustment cal m/c, manual adjust- recipe from central stor- load of recipe from cen- of recipe from central after download ment after download age, manual adjustment tral storage, manual ad- storage (one transacafter download justment after down- tion), no adjustment load after download Process Time Window / N2 Cabinet (Min / Max time control) Diferentiators: Data Collection, Data Validation, Decision Making no recording manual recording, man- automated recording, automated ual validation manual validation automated
[min max] Level 0 Level 1 Level 2 Level 3 Level 4 Statistical Process Control (SPC) Diferentiators: Data Collection, Data Validation, Decision Making (Lot Hold/ Tool Stop/ Trigger re-measurement) manual data collection, manual data collection, manual data collection, automated data collecmanual validation, man- manual validation, auto- automated validation, tion, automated validaual decision making mated decision making automated decision tion, automated deci(lot hold) making (lot hold) sion making (lot hold, tool stop, trigger remeasurement) Statistical Bin Analysis/ Automatic Lot Release (ALR) Diferentiators: Data Collection, Analysis level, Validation, Decision Making) manual input, h-bin manual input, h-bin manual input, h-bin analysis only, manual analysis only, auto analysis only, auto Defect Density Manage- Defect Density Manage- Defect Density Management System trigger, ment System trigger, ment System trigger, manual validation manual validation automated validation (lot hold) automated input (from test/handler summary), h-bin & s-bin analysis (ofline ALR), manual Defect Density Management System trigger, automated validation (lot hold) automated input (from test/handler summary), automated analysis of s-bin (ALR), automated Defect Density Management System trigger, automated validation (lot hold) Advanced Process Control/ Fault Detection and Classification Diferentiators: Tool Connectivity, Online Reaction, Out-of-Control Action Plan (OOCAP), Regular review process implemented tools not connected [no apc data flow] tool connected [apc data flow], some limits defined, e-mail notification 1st online reaction [tool stop, lot hold, inhibit next lot] has been established with oocap.
Level 0 Level 1 Level 2 Level 3 Level 4 Master Data Systems Availability Diferentiators: Standardization local, Data coverage (compared to overall site’s master data content), Timeliness non harmonized, no - limited local change - high level of standarduse of global master of global master data ization global/local, no data sets, only cascade sets, 4M (Man, Machine, local change of global from global to local on Method, Material) master data sets, 4M call partially available in (Man, Machine, Method, MES, batch/delayed Material) fully available cascade of global to in MES, global immedilocal ately cascade to local Master Data Systems Change/ Release Diferentiators: Maintenance [Manual / Auto], Release [Manual / Auto], Personal eficiency of the staf using the system, Capability of mass update automation, Workflow support (new workflow, workflow controlled data + performance management), Flexibility of data entry, Integrated efect analysis capability for change management, Analysis capability, Rollback capability manual maintenance/ semi-auto data changes synchronization/ en- from global planrichment, manual ning/product to MES release, manual data (non assisted) changes from global planning/product to MES, not connected data structures between global and local (tedious sync), analysis capability not set up, rollbacks are not supported auto mainte- semi-auto data changes nance/synchronization, from global planmanual release, mapped ning/product to MES data structures between (assisted), auto enrichglobal and local with ment of master data adaptions and aggre- locally high level of gation, ability to do analysis capability mass-change for global implemented change for non dependency items, ability to do mass-release for global change for non dependency items low level of analysis capability implemented, some manual enrichment of master data locally Master Data Static Systems Accuracy Diferentiators: Integrity [accuracy / timely] low data integrity, no in- formation on integrity available Master Data Dynamic Systems Accuracy Diferentiators: Integrity [accuracy / timely] low data integrity, no in- select/pick lists assisted formation on integrity data entry generated available, no aides (pick manual input lists) auto maintenance/synchronization, auto release, auto data changes from global planning/product to MES, equivalent data structures between global and local (fast sync), ability to do mass-changes for items of dependency, ability to do mass-release for global changes for items of dependency, full rollback capability on mass and individual changes full object dependent level of analysis capability implemented, not required enrichment of master data locally high data integrity, information on integrity available (plausibility check) measurable, highly consistent with static Master Data high data integrity, se- high data integrity, se- high data integrity, inlect/pick lists assisted lect/pick lists assisted formation on integrity data entry for all avail- data entry for reduced available (plausibility able selections selections (segment rel- check) measurable
evant) high data integrity, se- lect/pick lists assisted data entry generated from static Master Data Level 0 Level 1 Level 2 Level 3 Level 4 Operational Production Reporting Diferentiators: Standardization [Local, Global], Flexibility [Fix, flexible], Data Storage [Equipment, Local, Central], Integrity [accuracy / timely], Drill down functionality/capability, Automated report generation, Interlinking with mobile devices user generated reports by segment/site/user groups (Business Objects, excel), no standardization of reporting & manually generated, no link to mobile devices central generated cus- mixture of cluster - 1. regular report 1. on time tomized reports by seg- wide (FE & BE) and 2. cluster-wide harmo- 2. harmonized reports ment/site/user groups, customized reports nized reports between FE & BE no standardization on by segment/site/user a) same formula, data a) same formula, data cluster level - FE & BE, groups, no standard- source source no link to mobile de- ization between FE & b) same tool b) same tool vices BE, no link to mobile c) with diferent level of c) with diferent level of devices aggregation aggregation 3. no standardization be- 3. can be easily custween FE & BE tomized & automated 4. partially interlinking reporting to mobile devices 4. drill down functionality is available & easy to use 5. interface to manufacturing reporting 6. able to fulfill all levels of reporting from management to engineering 7. fully interlinking to mobile devices Aggregated Reporting Diferentiators: Standardization [Local, Global], Flexibility [Fix, flexible], Data Storage [Equipment, Local, Central], Integrity [accuracy / timely], Drill down functionality/capability, Automated report generation, Interlinking with mobile devices. user generated reports by segment/site/user groups (bo, excel), no standardization of reporting & manually generated, no link to mobile devices
Level 0 Level 1 Level 2 Level 3 Level 4 Data Analysis Diferentiators: Correlation along the Supply Chain, Usages of multiple relevant data sources/capability, Data completeness & availability, Data accuracy, Data integrity, Access speed, On-line data access, Stability, Handling of high volume data, Robustness & performance capability, Fast & interactive analysis capability/functionality, Coverage in terms of statistical methods (existing/available), Flexible to interact between diferent software system, Automation capability Data correlation along Data correlation within Data correlation within Data correlation be- Full Data correlation the supply chain is not FE or BE supply chain FE or BE supply chain tween FE & BE supply between FE & BE possible. Data availabil- is possible. Data avail- is possible. Data avail- chain is possible. Data supply chain. Full ity for certain facili- ability for all facilities ability for all facilities availability for all Data availability for ties along the supply within FE or BE sup- within FE or BE sup- facilities within FE & all facilities within chain. Data complete- ply chain. Data com- ply chain. Data com- BE supply chain. Data FE & BE supply chain ness & availability poor pleteness & availabil- pleteness & availabil- completeness & avail- (including relevant & not link to analysis ity moderate & partially ity good & linked to ability good & linked data from Silicon system. Low data accu- linked to analysis sys- analysis system. Moder- to analysis system. Foundry/Outsourcing racy with no monitor- tem. Low data accuracy ate data accuracy with Good data accuracy And Test (OSAT) with ing capability. No on- with manual monitor- manual monitoring ef- with semi-automated reference to contract). line access. Slow per- ing efort. No on-line ac- fort. Low on-line ac- monitoring. Partial Excellence data comformance of data access cess. Moderate perfor- cess. Good performance on-line access. Good pleteness & availability & unstable software so- mance of data access of data access & soft- performance of data & fully linked to analylution. Statistical meth- & software solution ful- ware solution fulfills for access & software sis system. Full on-line ods are not state of the iflls for simple analysis most of the analysis solution fulfills for all access. Excellence data art and not standard- tasks. Statistical meth- tasks. Statistical meth- of the analysis tasks. accuracy with fully ize within software so- ods are not state of the ods are state of the Statistical methods automated monitoring lution. Handling of high art and not standard- art and available in ex- are state of the art & reaction to deviations. volume data is not possi- ize within software so- isting non-harmonized and available in exist- Excellence performance ble. Ofline analysis soft- lution. Handling of high software solution. Tech- ing non-harmonized of data access & software is not aligned be- volume data is not possi- nology of software sys- software solution. ware solution fulfills for tween FE & BE. Inter- ble. Ofline analysis soft- tem is not state of the Technology of software all of the analysis tasks. action to other solution ware is not aligned be- art. Handling of high system is partially state Statistical methods are system is not possible. tween FE & BE. Inter- volume data is not possi- of the art. Handling of state of the art and No automation capabil- action to other solution ble. Ofline analysis soft- high volume data is par- within harmonized ity. system is not possible. ware is partially aligned tially possible. Ofline software solution.
No automation capabil- between FE & BE. Inter- analysis software is par- Technology of software ity. action to other solution tially aligned between system is state of the art.
system is partially possi- FE & BE. Interaction to Ability to handle high ble. Low automation ca- other solution system volume data according pability is partially possible. to requirement. Ofline
Moderate automation analysis software is capability. fully aligned across FE & BE. Full interaction to other solution system. Full automation capability.
Level 0 Level 1 Level 2 Level 3 Level 4 Storage & Retrieval System Diferentiators: Manual/ Assisted / Auto, Manual record/ Standalone / Link to MES, Link to Transportation System automated storage & re- automated storage trieval system, linked & retrieval system, to MES, no link to linked to MES, linked to transport system, (FE: transport system, (FE: stocker) stocker) Transport & Delivery System Diferentiators: Manual / Auto Transport System, Standalone / Link to Storage System, Link to Dispatching System, To Drop Point / Equipment, Link to Scheduling System automated transport automated transport system (conveyer, system, (conveyer, AGV, AGV), linked to MES, Automated Material linked to storage sys- Handling System), tem, to drop point / linked to MES, linked Equipment to storage system, to drop point / Equipment, linked to scheduling system Loading System [Robotics] Diferentiators: Manual / Auto Transport System, Standalone / Link to Storage System, Link to Dispatching System, To Drop Point / Equipment, Link to Scheduling System Level 0 Level 1 Level 2 Level 3 Level 4 Product (WIP)/ Device (Lot, Strip, Chip) Identification, Validation & Traceability Diferentiators: Level [Lot/ Device / Wafer / Strip / Chip], Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Shelf Life, Floor Life], Trace [Manual / Semi Auto / Auto], Equipment Internal wafer tracking manual identification automated identifi- automated identifi- Automated identifica- automated identifion lot level, manual cation on lot level , cation on lot level, tion on strip level, FE: cation on strip level, validation of in/out- manual validation of magazine, reel, automated identifica- automated identificaquantity in/out-quantity FE: automated identifi- tion on wafer level, tion on single device cation on wafer level, automated validation level after simulation, automated validation of in/out - quantity, equip- automated validation in/out-quantity ment internal wafer on strip and single tracking device level,
FE: chip level traceability [only applicable for some process steps] Material consumption & Wafer Material Identification, Validation & Traceability Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Shelf Life, Floor Life], Trace [Manual / Semi Auto / Auto] manual identification manual identification semi auto identification (sticker), manual record- (sticker), manual record- (barcode), semi auto ing (paper), manual val- ing (system), auto recording (barcode), idation (BOM, floor life, validation (BOM, floor auto validation (BOM, shelf life) life, shelf life) lfoor life, shelf life) semi auto identification automated identifi(barcode), automated cation (Equipment : recording (m/c reader), RFID/barcode), autoautomated validation mated recording (m/c (Equipment : BOM, reader), automated lfoor life, shelf life) validation (Equipment:
BOM, floor life, shelf life, consumption) Tool Identification, Validation & Traceability Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Life span, Maintenance cycle], Trace [Manual / Semi Auto / Auto] manual identification manual identification semi auto identification semi auto identification (sticker), manual record- (sticker), manual record- (barcode), semi auto (barcode), automated ing (paper), manual val- ing (system), automated recording (barcode), recording (m/c reader), idation (group, ID) validation (group, ID) automated validation automated validation
(group, ID) (group, ID) Carrier / Container Identification, Validation & Traceability Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Life span, Maintenance cycle], Trace [Manual / Semi Auto / Auto] same as above Equipment Identification & Validation Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Life span, Maintenance cycle], Trace [Manual / Semi Auto / Auto] same as above
Level 0 Level 1 Level 2 Level 3 Level 4 Operator Identification & Validation Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [qualify / not qualify], Trace [Manual / Semi Auto/ Auto] manual identification, manual recording (paper), manual validation (certification) manual identification, semi auto identification semi auto identification manual recording (sys- (manual login + bar- (single sign-on), autotem), auto validation code), semi auto record- mated recording, auto(certification) ing (barcode), auto vali- mated validation (certi
dation (certification) ifcation) Non Productive Products / Materials / Tools [durables] / Equipments Diferentiators: Identify [Manual / Semi Auto / Auto], Validate [Type, ID, Life span, Maintenance cycle, Floor life, Shelf Life], Trace [Manual / Semi Auto / Auto] manual identification manual identification semi auto identification (sticker), manual record- (sticker), manual record- (barcode), semi auto ing (paper), manual val- ing (system), auto recording (barcode), idation (BOM, floor life, validation (BOM, floor auto validation (BOM, shelf life) life, shelf life) lfoor life, shelf life) automated identification (Equipment: RFID) automated recording (Equipment : M/C reader) automated validation (certification) semi auto identification automated identifica(barcode), automated tion (eg : RFID/barcode), recording (M/C reader), automated recording automated validation (M/C reader), au(Equipment : BOM, tomated validation lfoor life, shelf life) (Equipment : BOM, lfoor life, shelf life, consumption) Unified Material Mapping Diferentiators: Scope [Full / partial supply chain] no identification standalone system, partial supply chain implementation, manual identification standalone system, partial supply chain implementation, auto identification linked with MES, par- linked with mes, full tial supply chain imple- supply chain implemenmentation, auto identifi- tation, auto identificacation tion Level 0 Level 1 Level 2 Level 3 Equipment Interface Diferentiators: No Connection / Serial / Ethernet, File Transfer/ SECS/GEM / Interface A no connection serial / GPIB / USB, file SECS/GEM - serial port, SECS/GEM - ethertransfer, legacy protocol (min 9600 baud rate - net (HSMS - high low data bandwidth) speed SECS messaging services, high data rate - 10mb/sec) SECS/GEM, Interface A (extreme high data rate - > 100mb/sec)
Equipment Data Diferentiators: Status [Up/Down], Event [Alarms / Start / Stop], Parameter [Input / Output], Result [Pass / fail], Frequency [Real time for the smallest Unit]
status - up/down status - signal from status - SECS/GEM, status - SECS/GEM, (tower light), event - equipment to external, event - unlimited alarm event - automated alarm (within equip- event - limited pre-set list from equipment, alarm list from equipment), result - complete list (manual selection), result - complete cy- ment, result - complete cycle/stop, parameter - result - complete cy- cle/stop, parameter cycle/stop, paraminternal view only cle/stop, parameter - in- - RMS capable, tool eter - RMS & APC ternal view only start/stop (input/output) capable, tool start/stop Automated Setup/Change Over Diferentiators: Triggering [Auto / Manual], Identification [Auto / Manual], Change over [Auto / manual] mechanism - manual, mechanism - manual, mechanism - auto mechanism - auto tool - manual, lot man- tool - manual, lot man- change by recipe con- change, tool - auto agement - no, recipe - agement - manual key trol, tool - manual, lot change, lot manageno in lot ID, recipe - man- management - scan ID, ment - by host control, ual recipe selection lot ID, recipe - RMS recipe - RMS auto
manual download download Equipment Health Monitoring Diferentiators: # of critical parameters to be monitored, Availability no monitoring, indica- monitoring [snap shot], monitoring [snap shot], tor / counter only only equipment status, equipment status & critsimple health moni- ical alarm, equipment toring on machine (eg: with intelligent sensor timeout: servo motor to provide local heath and communication monitoring - equipment within the equipment) related monitoring [snap shot], monitoring [real time], equipment status, criti- equipment status, critcal alarm & critical pa- ical alarm & critical rameter, real time APC, parameter, linked with health data from ma- lot ID, real time APC, chine used to have intel- health data from maligent process control - chine used to have intelofline and not real-time ligent process control (end of a day) ofline and real-time (every lot)
Level 0 Level 1 Level 2 Level 3 Level 4 Maintenance Diferentiators: Reactive, Proactive, Preventive, Predictive, Assisted Maintenance, Close Loop, Maintenance Monitoring run to fail [break down] time & volume based time & volume based time & volume based predictive modeling, maintenance, ifxed maintenance, ifxed maintenance, inte- automated scheduling schedule / volume schedule / volume, grated to SAP & based on production integrated to SAP, MES situation e.g. loading, equipment with intelli- integrated to SAP & gent sensor to provide MES local heath monitoring - equipment related, advice what needs to be changed before critical failure Input loading/Output loading (only backend) Diferentiators: Batch size, Validation Capability single input / single out- batch loading at input put loading, manual val- & output manual validaidation tion