Good question - just some weeks ago I summarized my personal method/approach on tech research: https://www.alp54.com/posts/awesome-tech-research/
Further getting in touch/playing around with technologies (prototyping, etc...) might help to get here as well.
"The best way to predict the future is to invent it." - Alan Kay
Reading the Twitter and blog posts of those who are is a second-best.
Historically, I seem to have an accuracy of about 60% in determining when something new on the scene will be important or not. Better than chance, but not by much.
Is your answer giving any insight to OP on how they should do it?