Robots are great tools but are incredibly limited to what they can do on their own. A human being working with a robot can do tremendously more than a robot (aka. machine) can do alone. That is not to say that we should not send robots ahead of human beings with a limited goal to set up some limited infrastructure. I could imagine sending a source of electric generation (not sure if nuclear is feasible without lots of water), a few habitat modules, a few storage modules, some vehicles, a few general purpose machine shops (for recycling, tools, some place to build stuff on the fly), and a few resource modules (containing raw materials that wouldnt be easily available locally).
The first generation of one way earth-martians are going to be the ones who need to figure out how to make Mars work for humans. Until human beings can exist on Mars without supplies from earth, Mars will be like Greenland was to the vikings in the 10th century or Antarctica to modern man. I mean, lets take a look at Antarctica. Its on the same planet, but we do not colonize it because it is so difficult to inhabit. Mars is colder than Antarctica. Mars is farther away than Antarctica. Mars does not have a breathable atmosphere like Antarctica. Mars does not have water like Antarctica. You can also come back from Antarctica. Antarctica is larger than the continental United States. Antarctica has LOTS of resources that is worth while for human beings to have. The gravity on Antarctica will not cause bone mass loss in human beings.
Worth while endeavor, but the real question is regarding the pay off.