It is widely agreed that coherent structures carrying a substantial share of the Reynolds stresses are important ingredients of free-shear turbulent flows, where they have led to applications such as improved control. The issue is less clear in attached wall-bounded flows, such as pipes, boundary layers and channels, whose mean velocity profiles lack known inertial modal instabilities.

However, the past quarter century has seen numerous proposals of simplified models for dominant structures in those cases, including exact travelling waves, orbits, linearised transients, and hairpin packets. Most of these structures are known to be important in transition, but less is known about whether they are present in fully developed wall-bounded turbulence in the same sense as in the free-shear case. Although it is all but certain that none of them can be found exactly in a fully turbulent flow, the interesting

question is whether some approximation can be found, approximation metrics can be defined, and quantitative information obtained about how close and how often a turbulent system comes to any of them. The purpose of this colloquium is be to bring together researchers working in the identification of simple models with the owners of turbulence data bases, to try to define what needs to be done. There are by now enough candidate structures, and enough temporally and spatially fully resolved DNS data bases, that such a program is beginning to be possible.