In this paper, we present an approach for the design of plant layouts in st
ochastic environments. We consider systems where the product mix and produc
t demand are subject to variability and where duplicates of the same depart
ment type may exist in the same facility. In contrast to a job shop layout,
we allow these duplicates to be placed in non-adjacent locations on the pl
ant floor and for flow allocation between pairs of individual departments t
o be made as a function of the layout and the product demand realization. W
e present a scenario-based procedure that iteratively solves for layout and
flow allocation. We show that having duplicates of the same departments, w
hich can be strategically located in different areas of the plant floor, ca
n significantly reduce material handling cost while effectively hedging aga
inst fluctuations in flow patterns and volumes. We show that the effect of
duplication is of the diminishing kind, with most of the cost reduction occ
urring with relatively few duplicates. We also show that the quality of the
obtained layouts can be quite insensitive to inaccuracies in estimating de
mand scenario probabilities.