Not theprof, but -- it varies between Universities, but the "panel" is always at least two. In some places you have to defend by giving essentially a lecture to both the academics and a more general audience, and answer questions. In others it's a quieter affair, in still other places it's somewhere in between I guess.
In all cases, I think the aim is the same: the purpose of the defence is to prove that you understand what you wrote in your thesis: both what it says and what it doesn't say. Almost by definition, if you've made it as far as the defence (or the "viva voce"), then the thesis should have the required content, subject to some level of corrections, to be awarded a pass. In particular this should be true because the candidate's supervisor should be competent enough, and engaged enough, to have ensured that the thesis was good enough to meet requirements.
There are always exceptions, but in general the candidates who are going to fail their defence either never made it that far to start with, because they dropped out, or turn out to have plagiarised their work. Everybody else should pass "with minor corrections", which usually means typos or minor clarifications; or, very occasionally, with "major corrections", which means that maybe the student will need to perform and discuss new experiments or write a new chapter on some theoretical point.