symmetric monoidal (∞,1)-category of spectra
The notion of globular operads is a variant of that of operads on which certain algebraic notions of higher category are based. The notion was introduced by Batanin; a globular operad is also called a Batanin operad.
A globular operad gives rise to a monad on the category of globular sets; one example is the free strict ω-category monad $T$ on globular sets. The monads which so arise may be characterized precisely as cartesian monads on globular sets over $T$ (itself a cartesian monad). This means that they are also examples of generalized multicategories relative to $T$.
A globular collection is a globular set $X$ equipped with a map
to $T(1)$, the underlying globular set of the free strict $\omega$-category on the terminal globular set. Hence the category $Coll$ is the slice category
The category of collections carries a monoidal product $\circ$ defined as follows. Given collections $f: X \to T(1)$, $g: Y \to T(1)$, the underlying globular set of $X \circ Y$ is given by pullback
and the requisite map $X \circ Y \to T(1)$ is given by the composite
where $\mu: T T \to T$ denotes multiplication of the monad $T$. The monoidal unit is the collection $u(1): 1 \to T(1)$ where $u: Id \to T$ is the unit of $T$, and the associativity and unit constraints may be defined by means of universal properties, taking advantage of the fact that $T$ is cartesian.
A globular operad is a monoid in the monoidal category $Coll$, (with the monoidal structure given by def. 2).
Each globular operad $f: P \to T(1)$, (as in def. 3), gives rise to a globular monad $M_P$ on $Set^{Glob^{op}}$. Abstractly, $M_P(X)$ is just the pullback
and the multiplication and unit for $M_P$ may be worked out from the multiplication and unit for the globular operad $P$.
A more concrete description of $M_P(X)$ may be worked out in terms of a concrete description of the free strict $\omega$-category $T(X)$. To describe this, first notice that every element $\tau$ of $T(1)$, which is essentially a pasting diagram built up out of globes of $1$, can be drawn as a globular set which we denote as $[\tau]$. The globes of $[\tau]$ are instances of globular cells as they appear in the pasting diagram $\tau$, and their sources and targets are then also instances of cells in $\tau$. (Batanin describes $T(1)$ in terms of trees, and the globular set $\tau$ is given formally in the tree language.)
Similarly, we can think of an element of $T(X)$ as a pasting diagram built out of globes in $X$, and such a pasting diagram can be thought of as having an underlying shape given by an pasting diagram $\tau$ in $T(1)$, together with a labeling of the pasting cells in $\tau$ by elements on $X$. The labeling is in fact just a morphism $[\tau] \to X$ of globular sets. Therefore we have an explicit formula for the set of $n$-cells of $T(X)$:
and similarly, for a globular operad with underlying collection $f: P \to T(1)$,
The category of operators of a globular operad $A$ is (the syntactic category of) a homogeneous globular theory $i_A \colon \Theta_0 \to \Theta_A$ and every globular operad is characterized by its globular theory. See there for more details
Write $\omega$ for the globular operad whose category of operators, see above, is the Theta category $\Theta$.
The category $Str\omega Cat$ of strict ω-categories is equivalent to that of algebras over the terminal globular operad. Hence it is the full subcategory of that of ω-graphs which satisfy the Segal condition with respect to the canonical inclusion $\Theta_0 \to \Theta$ that defines its globular theory: we have a pullback
As a refinement of the above example:
In the Batanin (or Leinster) theory of $\infty$-categories, there is a universal contractible globular operad $f: K \to T(1)$, where each element $x \in K(n)$ is thought of as a way of (weakly) pasting together the underlying shape $f(x)$. The contractibility implies that for every two different ways of pasting together the same shape, i.e., two elements $x, y \in K(n)$ such that $f(x) = f(y)$ and such that $x$ and $y$ have the same source and have the same target, there is an $(n+1)$-cell in $K(n+1)$ mediating between them, with source $x$ and target $y$, and which maps to the identity $(n+1)$-cell on $f(x)$.
A Batanin ω-category is a globular set with a $K$-algebra structure.
A review and characterization in terms of globular theories is in section 1 of
Other work on globular operads :