The general public has an image of academia as a society protected from the real world where people sit around and think, experiment, and create new solutions to every problem that exists in the world. Though this view is partially accurate, universities are in fact strongly tied to the real world and subject to several major aspects of the real world, including:
Generally, no research is done without funding. Funding pays for graduate students, equipment, events, marketing, travel, and anything else related to a research project except the professors' salaries (in the United States; I hear it's different in Europe), which generally come straight from the university. Each professor typically has a couple of grants on which they are the main person (PI, or principal investigator), and a few other grants from which they also get some money, but on which they're not the PI.
A PhD student is a huge investment for a professor; a PhD student can cost more than $100,000 per year, including stipend, tuition, office space, travel expenses, and computers and other equipment. [Stanford CS Redbook] Understandably, professors are hesitant to take a new student into their group without some confidence that she will be a worthwhile investment for them! Professors spend a lot of time writing grant proposals and marketing their research to keep the money coming in, since all grants are finite.
Most people think of professors as highly intellectual, pipe-smoking, bearded males who have thought everything through very carefully with great wisdom before they do anything. And people think of graduate students as people who aspire to be this non-human breed of superintellectuals. In reality, professors are humans too. In general, academics are those who have thought very carefully about their particular areas of interest, both academic and non-academic; but, like all human beings, this doesn't mean they have thought more carefully about their actions, morals, motivations, and behavior than other reasonably thoughtful human beings. Unlike the ivory tower image most people have of academia and graduate school, you will find that most meetings never start on time, that a professor's enthusiasm about a student's work can spike and fizzle based on seemingly inconsequential things such as a nice-looking plot, and that shows of excitement about you coming to their graduate school are not an absolute guarantee that they will take you on as a student. Professors operate under the constraints of human nature and finite funding, so they will naturally be more hesitant to actually take on new students than what they may seem like at each school's admit weekend, where their job is partly to advertise their university and their research to you.
You will also see that, similar to industry and the rest of the world, a lot of success hinges on who you know. Networking is key, and professors at the top schools are generally very well connected. Even getting into top graduate schools depends heavily on what your undergraduate professors tell professors at those schools about you. Just imagine, a PhD program may get 1000 applications, and out of that they can only pick 50 to be admitted, even though, say 200 of those applicants are qualified to be in any top school. Professors will naturally pay much more attention to the applications of students their colleagues have said good things about, than someone with a couple of publications whom none of their friends mentioned.
Most PhD students I have talked to came into graduate school with a dream of making amazing contributions to their field, then coming up with a research plan after a year or two, and then taking lots of shortcuts and turns to finish up each project and their whole PhD. A PhD alone is a lot of work, but is typically a fairly small (though significant in some way) contribution to a field. As I said before though, don't get discouraged if your work does not live up to your expectations. The goal of research is to achieve our dreams of great contributions to society and knowledge. However, unlike what many students think coming into graduate school, these great contributions rarely occur from the efforts of a single student during a fraction of their PhD. Typically, great research is done in small steps: first an idea, then testing, multiple failed attempts at solutions that increase your knowledge about a problem, and eventually a solution that may be better than what has ever existed before. And though a miniature version of this happens during most PhDs, most great contributions in research occur over a much longer time span, and with a lot more manpower, than a lone student's PhD work, or even a whole group's work in a short period of time. When I say great research, I mean truly groundbreaking stuff that breaks paradigms and solves major problems, not just well-done incremental work.
Fortunately, the research community understands this, and so we typically can publish any small steps of progress we make, as long as what we have done is good enough to be considered at least a dent in the field. So typically research is done in these small steps, with conference and/or journal papers marking each of these steps and making them known to the world. So researchers within a field will typically call a well written paper "great research," even though society as a whole may not consider that paper to matter much at all. This somewhat awkward and somewhat funny trade-off between idealism (wanting to do great research) and reality (the fact that research is hard, slow, and is most productively done in small steps) will characterize a lot of your PhD work and research career as a whole. Also, you will probably find that what society really wants (a comprehensive, 100%-reliable solution to a problem) and what your individual research leads to (partial solutions to major problems, or comprehensive solutions to small problems) are at odds with each other.
Unlike math research, which I was more familiar with before coming to graduate school, engineering research is focused mainly on producing and building systems and solutions, and less on great ideas and deep thinking. The typical student in an engineering lab in a top school will spend a lot of their time building things or writing code, tweaking things, and trying to understand the system they or others have built, to get results that seem to be good enough based on a fairly high standard.
But this is only about half of the work. 50% or more of the work involves presenting this work to your colleagues, your professor, your lab, audiences at conferences, and reviewers at conferences and journals. The exact amount of emphasis placed on the quality of papers and presentations vary widely between professors and labs, but generally speaking, the amount of work that goes into presenting and selling your work is comparable to how much work you put into actually building something and getting results.
Some engineering labs are more theory- and math-driven than others that are more results- and system-building-driven, but typically even the theoretical labs often operate like engineering labs and not math labs. Building a decent system to attack a problem is typically regarded more favorably than a great idea. Even though great ideas are what seed some of the greatest achievements, only a finished accomplishment that is published or highly recognized by others will get the social recognition and impact needed to spark more collaboration and get more funding to continue doing research on a particular problem. And such accomplishments are the only tangible things that lead to a PhD thesis.