IMO:
R&D is all about specialization, so choose an area you'd like to work in. You don't need to choose immediately, you can play with various stuff to find the right kind of work for you (the factors that matter for me are: the amount of low-level/high-level coding, the heaviness of math, the size of projects).<p>As for PhD, I think that, as long as you're exceptional in your area of expertise, you can get R&D jobs without it. On the other hand, getting to that level can easily take as long or longer than doing a PhD. Also, I think you can probably gradually advance into more researchy positions through smart career progression, starting with more development-oriented jobs in R&D areas combined with doing plenty of research on your own.
Right out of college I found my place at a startup working on R&D. I was the only engineer/technical person there so it was a great experience for me. I was given free reign over product development so out really helped me grow by forcing my to learn every aspect of the development process of going from idea conception, to prototype, to running a full swuite of tests, determining component reliability and life cycle, finding manufacturers/suppliers, and eventual winding up with a reliable, efficient, and cost effective product. After that job I could have undoubtedly gone on to work in R&D in the corporate world, but I decided to go into management for various reasons not related to work.<p>So my advice to you is see if you can get in on the ground floor at either a startup or small company that is in R&D and you'll be forced to learn a ton of provable marketable skills that will allow you to move on and work for a larger company without a PhD.