If it is not patented, it is legal to use it (in most countries). Some (very big) companies have a lot of people working on reading research papers to sort out if something profitable can be done based on them.
However, most of the time it is cheaper to contract with the university/researcher in order to really understand how things are really done (most papers are very elliptic).
In most of the cases you will have to do a lot of work yourself, in order to successfully apply some research results.
Typically, there's a big difference between theoretical and applied research.
Applied research is more focused towards implementation problems, realizability, etc. and can be more easily used to build a product/service. In most of the cases, such research is accompanied by a patent.
On the other hand, theoretical research mostly involves the development of solutions to currently open problems. It leaves out the implementation part; at best you will see a simple simulation example with no practical significance.
I would say that in most cases you will need to develop your own "research" results and extend existing ones, that will lead to a product.
Patents aren't the only concern. When dealing with written ideas think COPYRIGHT first. You need to ask the copyright holder in order to legally republish the research publication. You can cite and paraphrase freely, of course.