Maintenance pooling to maximize fab uptime

Maximizing process tool uptime has remained a core challenge for the manufacturing of advanced semiconductors over time. This becomes more important than ever with newest process technologies and the scale of 450mm fabs that are under consideration, driving an ever increasing value of each wafer. One key step towards this goal is to synchronize the timing of scheduled maintenance of peripheral equipment and critical sub-systems such as vacuum pumps and abatement (exhaust management) systems with that of the process tools they support. Transitioning from calendar-based to a knowledge-based solution is dependent upon increased intelligence within the critical sub-systems and concomitant improvements in communications between them and the process tool. This demands a significant degree of intelligence in the peripheral equipment itself so that an accurate Remaining Useful Life (RUL) indication can be realized, and used to synchronize maintenance activities with the maintenance schedule of the process tool. Effective knowledge sharing direct to the process tool and/or via the factory information and control system (FICS) is required, allowing the peripheral equipment to interactively calculate RUL, assess the risk of operation beyond routine maintenance intervals and to potentially suggest appropriate actions to minimize this risk. Robust risk analyses can be performed to determine if RUL of critical sub-systems can be managed in such a way to extend operation until the next planned process tool PM, enabling a later synchronized maintenance activity. In the absence of alternative solutions the least risk decision is to schedule maintenance activity and absorb the implied cost if a fault is predicted to occur before the next tool PM. Together, these approaches empower the tool owner to make evidence based judgments of both the pro's and con's of synchronization with other maintenance procedures. This paper discusses how such synchronization might be achieved, and the risks and considerations required for a successful implementation.