We analyze the value of mobile production capacity and transshipment in a
supply chain with geographically distributed production facilities and
epoch-dependent demands. We describe the L location, Y mobile production unit
problem as a problem of sequential decision making under uncertainty to
determine transshipment, mobile production capacity relocation, and
replenishment decisions at each decision epoch. We model a data-driven demand
forecasting capability based on the existence of a partially observed
stochastic process, the modulation process, that affects demand, is unaffected
by the actions of the decision maker, and reflects the influence of exogenous
and partially observed forces (e.g., the macro economy) on decision making
environments. We model a specially structured partially observed Markov
decision process, develop several heuristic policies, and compare them
computationally, demonstrating that two decision making approaches, a
centralized approach and a decentralized approach, show considerable promise
for producing high quality heuristics. We show for an instance set with five
locations that production capacity mobility and transshipment can improve
systems performance by as much as 41% on average over the no flexibility case
and that production capacity mobility can yield as much as 10% more savings
compared to transshipment capability when present as the only form of
flexibility.