I gave my talk at the Maastricht workshop yesterday and so here are the slides (Maastricht talk).

Matroid theory is not especially well known (in fact, when he gave a talk here, Brian Alspach used a result from matroid theory and described the subject as “the one that is allocated the smallest and furthest away room in any conference that has parallel sessions”) and so usually it’s necessary to put a lot of basic definitions into any talk. Fortunately this occasion, where everyone is working in matroids, is an exception and so I could launch straight in.

In some sense, binary matroids are a generalization of graphs, and so any question that can be asked about graphs can be asked unchanged about binary matroids. In the 1930s Klaus Wagner proved the famous excluded minor result

A graph is planar if and only if it does not have the complete graph $K_5$ or the complete bipartite graph $K_{3,3}$ as a minor.

He also gave a description of the graphs that do not have $K_5$  as a minor, and similarly the class obtained by excluding $K_{3,3}$.

We (that is,  me and my friends from Wellington NZ, Geoff Whittle and Dillon Mayhew) are trying to extend these two results to binary matroids and get descriptions of the exact structure of those classes. It turned out that the case for excluding $K_{3,3}$ was tractable, though we ended up with a long 113 page paper that has just appeared in the Memoirs of the American Mathematical Society. But excluding $K_5$ seems much more difficult and we really don’t know where to go from here.

My talk basically describes what we’ve done and where we’re stuck!