Kan Complexes can be seen as a generalization of groupoids, mostly called (weak) infinity groupoids in this context.
On groupoids we can define the \textbf{group of bisections} the following way:
Let $G \Rightarrow M$ be a groupoid with arrow set $G$, object set $M$,
target $d_0: G \to M$,
source $d_1: G \to M$, 'units' $s_0: M \to G$ and product
$*:G \times_{(d_0\times d_1)} G \to G$
Then a bisection is a map $\sigma: M \to G$ right inverse to $d_1: G \to M$, such that $d_0\circ b : M \to M$ is bijective.
(Of course this would work similar if we say b is right inverse to $d_0$ and $d_1\circ b$ is a bijection)
The set of bisections has a group structure, with product defined on elements
of M by
$$(b_1 \star b_2)(m) := b_1(d_0\circ b_2(m))*b_2(m)$$
and unit s_0 .
Now, the question is, if this can be generalized or reformulated to something on oo-groupoids, replacing the associative product by something else.
