I finished my PhD in probability a few years ago and left to go work in the industry as software engineer in operating systems and databases. I sometimes ponder whether I can go back to doing research, perhaps on a part-time basis, if I somehow transform my character to have more self-discipline. When I was an undergraduate student, I really loved mathematics. Even now, on my spare time, I frequently go over my old courses and, as Terrence Tao puts it, learn and re-learn old materials. Each time it brings me great pleasure and newer insight.
What I have trouble with is actually spending time solving research problem. With absolutely no intention of offending mathematicians, I find that most of the open problems in mathematics don't hold that much interest to me because I find them too narrow. I really had a tough time with the proofs in my PhD thesis because they are hard and I couldn't sustain consistent work each day. I prefer discovering new connections and asking my own questions, such as pondering on the connection between complex analysis and probability or asking "what if I remove this assumption" and see if I can derive new theorems. Even then, I usually can only do that for one day or two. I simply cannot do it day after day like a professional researcher. I would go back to my old habit of debugging and writing programs which I can do for 8 hours a day five days a week no problem.
Thank you for reading so far. My overall question is, how do I determine if my lack of success in research mathematics is due to lack of interest in research or lack of self-discipline? Can I just use something like the pomodoro technique and build up my tolerance for prolonged mathematics research?