A military commander has to allocate resources to different courses of action in such a way as to achieve some desired set of outcomes. His view of what can be achieved by the different actions will vary with time as he sees the outcomes of his earlier choices. The outcomes also have intrinsic uncertainty. How can a mathematical model be made of this? At a different level, a defence organisation faces a corresponding problem when it has to partition funding between the armed services and defence technology research and development. The Study Group focused on the latter question, showed how it can be formulated as a problem in optimal control with delay, and studied examples of its behaviour numerically.

Problem presented by
Tim Gardener and Richard Hadji, Dstl

Study Group contributors
Chris Breward (University of Oxford)
Huaxiong Huang (York University, Canada)
Andrew Lacey (Heriot-Watt University)
Steve Noble (Brunel University)
John Ockendon (University of Oxford)
Ashley Pitcher (University of Oxford)
Eddie Wilson (University of Bristol)
Tom Witelski (Duke University, USA)