I will discuss the results of recent work applying a nonlinear optimisation technique to the shear flow transition problem.
This technique extends now well-known (linear) transient growth analysis to finite amplitude with the aim of detecting the `best' way - the minimal seed - to disturb a system out of the basic (laminar) state's basin of attraction (see Kerswell, Pringle & Willis, Rep. Prog. Phys. 77 085901, 2014 for a review). An example will be given of how this new capability can be used to design more nonlinearly stable flows and then possible future directions will be outlined.