Eniac-improve
Structure de mise en forme 2 colonnes

WP3

Considering the huge amount of capital expenditure represented by the process tools in a leading edge semiconductor facility, it is absolutely critical to maximize the use of these assets. To improve the current availability and reliability of the production tools, it is necessary to move beyond the present APC and FDC approaches based on real time control of the tools’ behaviour, towards a predictive approach. The overall objective of WP3 is to develop a Predictive Equipment Behaviour system to support this new approach.

The prediction of equipment behaviour will allow to anticipate failures and to launch ad-hoc preventive maintenance actions. It is the basis to set up a Predictive Maintenance strategy and defines the main objectives of WP3:

  • Improvement of effectiveness by describing the deviation of the actual equipment behaviour from the status of normal productive operation based on the relevant key parameters of the equipment (“health monitoring”)

  • Decrease of equipment failure rate (unscheduled down time) by detecting out of control deviations before the equipment is going unscheduled down enabling appropriate  containment actions to avoid this final status

  • Speedup of equipment recovery by classifying out of control deviations and identifying its root cause enabling fast determination of the corrective actions required to bring the  equipment back to the status of normal productive operation

 

 

To achieve these objectives, the work in WP3 is organized into 5 major tasks comprising 18 main and numerous sub-deliverables.

 

 

The generic workflow within WP3 follows an iterative 3-step-approach: data sets built from currently accessible data extracted from equipment, sensors and information systems (1), will be used to construct predictive models based on physical and statistical approaches (2) that will then be tested and assessed with new data sets (3). Based on these results, a more accurate version of data sets will be defined (5) and the models will be refined (4) until good prediction accuracy and model stability is reached.

 

 

The work structure of WP3 is defined by the assigned collaboration of 21 of the 35 Improve beneficiaries organized within 3 clusters: 9 Semiconductor Manufactures (SCM), 6 Solution Providers (SP), 3 Institutes and 3 Universities (I&U).

 

 

As WP3 is part of the core of the Improve work plan, mutual interrelations are defined to all other work packages within the project.

To top