To be productive and profitable in a modern semiconductor fabrication envir
onment, large amounts of manufacturing data must be collected, analyzed, an
d maintained. This includes data collected from in- and off-line wafer insp
ection systems and from the process equipment itself. This data is increasi
ngly being used to design new processes, control and maintain tools, and to
provide the information needed for rapid yield learning and prediction. Be
cause of increasing device complexity, the amount of data being generated i
s outstripping the yield engineer's ability to effectively monitor and corr
ect unexpected trends and excursions. The 1997 SIA National Technology Road
map for Semiconductors highlights a need to address these issues through "a
utomated data reduction algorithms to source defects from multiple data sou
rces and to reduce defect sourcing time." SEMATECH and the Oak Ridge Nation
al Laboratory have been developing new strategies and technologies for prov
iding the yield engineer with higher levels of assisted data reduction for
the purpose of automated yield analysis. In this article, we will discuss t
he current state of the art and trends in yield management automation. (C)
1999 American Vacuum Society. [S0734-2101(99)04404-8].