Let $X$ be a compact Hausdorff space.Put $Vec(X)$, the space of all real (or complex) vector bundles over $X$.We put also $Vec_g(X)$, the space of all Riemannian vector bundles over $X$, that is the space of all vector bundles equiped with a continuous fiber wise inner product. By imitation of three equivalent relations defined on $Vec(X)$ in the theory of topological $K$-theory, we define a metric version of the same equivalent relations as follows:
1)Two Riemannian vector bundles $(E,g)$ and $(F,h)$ are equivalent if there is isometric bundle isomorphism between them.
2)Two Riemannian vector bundles $(E,g)$ and $(F,h)$ are equivalent if there is a trivial bundle $\epsilon_k$ with the obvious metric structure such that $E\oplus \epsilon_k$ is equivalent to $F \oplus \epsilon_k$ in the sense of 1) where each direct sum bundle is equipped with direct sum.metric
3)Two Riemannian vector bundles $(E,g)$ and $(F,h)$ are equivalent if there are two trivial bundles $\epsilon_m$ and $\epsilon_n$ with obvious metrics such that $E\oplus \epsilon_m$ is equivalent to $F \oplus \epsilon_n$ in the sense of 1) where each direct sum bundle is equipped with direct sum.metric.
Does this idea produce a new kind of $K$_theory? Does it introduce a functor from the category of compact Hausdorff space to the category of Groups? Can we extend this functor from the category of commutative unital $C^*$ algebras to the category of non commutative unital $C^*$ algebras.