I have argued elsewhere that this notion of a global metric for morality is doomed https://philosophy.stackexchange.com/a/26553/9166. But, being a hypocrite, I still have a favorite one, which I feel is physicalistically motivated.
Any vote for a physicalist version of utility could not really be tied to human pleasure. We do not play any special role in physics, so there does not seem to be a basis for any physics-based system that provides us in particular with a special role. A genuine distinction requires a difference.
You could argue for a utility that marks a genuine difference between the living and the non-living, but living does not require a sense of pleasure. Amoebas live. One doubts they suffer, the Buddha not withstanding.
So among singular criteria for ethics that have been prominently proposed, we are reduced to a range from Schopenhauer to Nietzsche. Utility has to be in survival value or in the overall ability to shape one's own circumstances. To my mind, what these have in common relative to physics, is that they are about information. Survival involves creating genetic copies that maintain the information already here, power involves inserting information that is specific to you into the environment.
I would vote for this as an approach to utility, and adopt Terrence McKenna's 'Theory of the cosmic giggle'. What is good is what overall produces the greatest opportunity for novelty. To me this suggests a sort of quanititative, evolutionary Kantianism. As the process that produced us, evolution, not just the biological form, but the overall process of elaborating information, is in itself good. It is not just our cause, it is what we are here for.
Autonomy should be respected not as some force of truth, but as the conduit for beings to attribute novelty to the world. And that means that it is not an absolute inviolable quality -- but other than that, I think Kant has the basic principles right.
We cannot know what kind of autonomy will lead future evolution, so destroying or unnecessarily directing it is an unnecessary risk. At the same time, we are made of autonomy, and limiting ourselves pointlessly is an equal risk. That leads into something like the means-ends version of the Categorical Imperative, as the measure of utility that makes for a world where information is served. More equality is better, more variety is better, more complexity is better, as long as it is natural. Survival and a certain level of comfort is necessary to contribute, among members of most species, so approximate rule-utilitarianism, with a bias to weigh pain over pleasure, is a vague corollary.
I think the inborn human concern for fairness in social competition is not an idiosyncratic attribute we have as a primate species with given social goals. I mean, it is, but it also expresses something more general, and we can rely upon that instinct to point out a more general principle of which it is basically a local variant.